Chapter 1 Introduction to Machine Learning “Computers are able to see, hear and learn. Welcome to the future.” — Dave Waters Machine Learning (ML) isa promising and flourishing field. It can enable top management ofan organization to extract the knowledge from the data stored in various archives of the business organizations to facilitate decision making. Such decisions can be useful for organizations to design new products, improve business processes, and to develop decision support systems. + Explore the basics of machine learning + Introduce types of machine learning j + Provide an overview of machine learning tasks + State the components of the machine learning algorithm, + Explore the machine learning process * Survey some machine learning applications 1.1 NEED FOR MACHINE LEARNING Business organizations use huge amount of data for their daily activities. Earlier, the full potential of this data was not utilized due to two reasons. One reason was data being scattered across different archive systems and organizations not being able to integrate these sources fully. Secondly, the lack of awareness about software tools that could help to unearth the useful information from data. Not anymore! Business organizations have now started to use the latest technology, machine learning, for this purpose. Machine learning has become so popular because of three reasons: 1. High volume of available data to manage: Big companies such as Facebook, Twitter, and YouTube generate huge amount of data that grows at a phenomenal rate. It is estimated that the data approximately gets doubled every year.eer eicran Introduction to Machine Learning “Computers are able to see, hear and learn. Welcome to the future.” — Dave Waters = Machine Learning (ML) isa promising and flourishing field. It can enable top management of an organization to extract the knowledge from the data stored in various archives of the business organizations to facilitate decision making. Such decisions can be useful for organizations to design new products, improve business processes, and to develop decision support systems. ‘+ Explore the basics of machine learning * Introduce types of machine learning * Provide an overview of machine learning tasks + State the components of the machine learning algorithm + Explore the machine learning process * Survey some machine learning applications 1.1 NEED FOR MACHINE LEARNING Business organizations use huge amount of data for their daily activities. Earlier, the full potential of this data was not utilized due to two reasons. One reason was data being scattered across different archive systems and organizations not being able to integrate these sources fully. Secondly, the lack of awareness about software tools that could help to unearth the useful information from data, Not anymore! Business organizations have now started to use the latest technology, machine learning, for this purpose. Machine learning has become so popular because of three reasons; 1. High volume of available data to manage: Big companies such as Facebook, Twitter, and YouTube generate huge amount of data that grows at a phenomenal rate, It is estimated that the data approximately gets doubled every year. 42» Machine Learning ‘The hardware cost has also dro et OPy distribute, and transmit the ge st of storage has reduced. digit : d reason is that the co: eee process, store, Therefore, it is easier now to capture, information. sue ; / .3, Third reason for popularity of machine learning is the availabilty otoon seo j now. Especially with the advent of deep learning, many alg lable fo, machine learning. : ; ; With the popularity and ready adaption of machine learning by business organization, it has become a dominant technology trend now. Before starting the a a cae % journey, let tus establish these terms - data, information, knowledge, intelligence, 2 - A knowledge pyramid is shown in Figure 1.1. Intelligence (applied knowledge) Knowledge (condensed information) Information (processed data) Data (mostly available as raw facts and symbols) Figure 1.1: The Knowledge Pyramid What is data? All facts are data. Data can be numbers or text that can be processed by a computer. Today, organizations are accumulating vast and growing amounts of data with data sources such as flat files, databases, or data warehouses in different storage formats. j Processed data is called information. This includes patterns, associations, or relationships f among data. For example, sales data can be analyzed to extract information like which is the fast selling product. Condensed information is called knowledge. For example, the historical patterns and future trends obtained in the above sales data can be called knowledge. Unless knowledge is extracted, data is of no use. Similarly, knowledge is not useful unless it is put into action. Intelligence is the applied knowledge for actions, An actionable form of knowledge is called intelligence. Computer systems have been successful till this stage. The ultimate i objective of knowledge pyramid is wisdom that represents the maturity of mind that is, 50 f@t exhibited only by humans. 7 Here comes the need for machine learning. The objective of machine learning is to proces these archival data for organizations to take better decisions to design new products, improve th® business processes, and to develop effective decision support systems,Introduction to Machine Learning «3 1.2 MACHINE LEARNING EXPLAINED Machine learning is an important sub-branch of Artificial Intelligence (Al). A frequently quoted definition of machine learning was by Arthur Samuel, one of the pioneers of Artificial Intelligence. He stated that “Machine learning is the field of study that gives the computers ability lo learn without being explicitly programmed.” The key to this definition is that the systems should learn by itself without explicit programming, How is it possible? It is widely known that to perform a computation, one needs to write programs that teach the computers how to do that computation. In conventional programming, after understanding the problem, a detailed design of the program such as a flowchart or an algorithm needs to be created and converted into programs using a suitable programming language. This approach could be difficult for many real-world problems such as puzzles, games, and complex image recognition applications. Initially, artificial intelligence aims to understand these problems and develop general purpose rules manually. Then, these rules are formulated into logic and implemented in a program to create intelligent systems. This idea of developing intelligent systems by using logic and reasoning by converting an expert’s knowledge into a set of rules and programs is called an expert system. An expert system like MYCIN was designed for medical diagnosis after converting the expert knowledge of many doctors into a system. However, this approach did not progress much as programs lacked real intelligence. The word MYCIN is derived from the fact that most of the antibiotics’ names end with ‘mycin’. The above approach was impractical in many domains as programs still depended on human expertise and hence did not truly exhibit intelligence. Then, the momentum shifted to machine learning in the form of data driven systems. The focus of Al is to develop intelligent systems by using data-driven approach, where data is used as an input to develop intelligent models. The models can then be used to predict new inputs. Thus, the aim of machine learning is to learn a model or set of rules from the given dataset automatically so that it can predict the unknown data correctly. . As humans take decisions based on an experience, computers make models based on extracted patterns in the input data and then use these data-filled models for prediction and to take decisions. For computers, the learnt model is equivalent to human experience. This is shown in Figure 1.2. Experience Decisions ——+| Humans |__ (a) Data Model Data- Learning program |» base () Figure 1.2: (a) A Learning System for Humans (b) A Learning System “ for Machine Learning Often, the quality of data determines the quality of experience and, therefore, the quality of the learning system. In statistical learning, the relationship between the input x and output y isMachine Learning ce xe, fis the learning function that maps the inp | aspect of forming 2 model in statistical leaming of input to output. ta in a model. Formally stated, a model ig 9) 4 modeled as a function in the form y = f(x). He! to output y. Learning of function fis the crucial In machine learning, this is simply called mapping The learning program summarizes the raw dal « explicit description of patterns within the data in the form of: 1. Mathematical equation 2. Relational diagrams like trees/graphs 3. Logical iffelse rules, ot 4, Groupings called clusters : In summary, a model can be a formula, procedure or representation that can generate dat decisions, The difference between pattern and model is thatthe former is local and applicable only to certain attributes but the latter is global and fits the entire dataset. For example, a model can be helpful to examine whether a given email is spam or not. The point is that the model is generated automatically from the given data. Another pioneer of Al, Tom Mitchell's definition of machine learning states that, “A computer E, with respect to task T and some performance measure P, program is said to learn from experience if ts performance on T measured by P improves with experience E.” The important components of this definition are experience E, task T, and performance measure P. For example, the task T could be detecting an object in an knowledge of object using training dataset of thousands of images. This is called experience E i So, the focus is to use this experience E for this task of object detection T. The ability of the system | to detect the object is measured by performance measures like precision and recall. Based on the performance measures, course correction can be done to improve the performance of the system. | Models of computer systems are equivalent to human experience. Experience is based on data. 4 Humans gain experience by various means. They gain knowledge by rote learning. They observe others and imitate it. Humans gaina lot of knowledge from teachers and books. We learn many things 1 by trial and error. Once the knowledge is gained, when a new problem is encountered, humans ! search for similar past situations and then formulate the heuristics and use that for prediction. But, in systems, experience is gathered by these steps: 1. Collection of data 2. Once data is gathered, abstract concepts are formed out of that data. Abstraction is used to generate concepts. This is equivalent to humans’ idea of objects, for example, we have some idea about how an elephant looks like. image. The machine can gain the 3. Generalization converts the abstraction into an actionable form of intelligence. Tt can be viewed as ordering of all possible concepts. So, generalization involves ranking of concepts, inferencing from them and formation of heuristics, an actionable aspect of intelligence. Heuristics are educated guesses for all tasks. For example, if one runs OF i encounters a danger, it is the resultant of human experience or his heuristics formation. | In machines, it happens the same way. 4, Heuristics normally works! But, occasionally, it may fail too. It is not the fault of heuristics as it is just a ‘rule of thumb’, The course correction is done by taking evaluation measures. Evaluation checks the thoroughness of the models and to-d° course correction, if necessary, to generate better formulations. ifIntroduction to Machine Learning © 5 4.3 MACHINE LEARNING IN RELATION TO OTHER FIELDS Machine learning uses the concepts of Artificial Intelligence, Data Science, and Statistics primarily. Itis the resultant of combined ideas of diverse fields. 1.3.1 Machine Learning and Artificial Intelligence Machine learning is an important branch of AI, which is a much broader subject. The aim of Al is to develop intelligent agents. An agent can be a robot, humans, or any autonomous systems. Initially, the idea of AI was ambitious, that is, to develop intelligent systems like human beings. The focus was on logic and logical inferences. It had seen many ups and downs. These down periods were called AI winters. The resurgence in AI happened due to development of data driven systems. The aim is to find relations and regularities present in the data. Machine learning is the subbranch of Al, whose aim is to extract the patterns for prediction. It is a broad field that includes learning from examples and other areas like reinforcement learning. The relationship of Al and machine learning is shown in Figure 1.3. The model can take an unknown instance and generate results. “Artificial intelligence Machine learning Figure 1.3: Relationship of Al with Machine Learning Deep learning is a subbranch of machine learning. In deep learning, the models are constructed sural network technology. Neural networks are based on the human neuron models. Many using ne B 5 in functions that trigger further neurons to neurons form a network connected with the activatio perform tasks. 1.3.2 Machine Learning, Data Science, Data Mining, and Data Analytics Data science is an ‘Umbrella’ term that encompasses many fields. Machine learning starts with data, Therefore, data science and machine learning are interlinked. Machine learning is a branch of data science. Data science deals with gathering of data for analysis. It is a broad field that includes:6 Blane Bean Big Data Data science concerns about collection of data, Big data is a field of data sciencg that deals with data’s following characteristics: Volume: Huge amount of data is generated by big companies like Facebook, Twitter, YouTube. Data is available in variety of forms like images, videos, and in different formats, t refers to the speed at which the data is generated and processed. Big data is used by many machine learning algorithms for applications such as language trans. lation and image recognition. Big data influences the growth of subjects like Deep learning. Deg Tearning is a branch of machine learning that deals with constructing models using neural networks, . Variet 2 . Velocity Data Mining Data mining’s original genesis is in the business. Like while mining the earth ong gets into precious resources, itis often believed that unearthing of the data produces hidden infor. mation that otherwise would have eluded the attention of the management. Nowadays, many consider that data mining and machine learning are same. There is no difference between these fields except that data mining aims to extract the hidden patterns that are present in the data, whereas, machine learning aims to use it for prediction. Data Analytics Another branch of data science is data analytics. It aims to extract useful knowledge from crude data. There are different types of analytics. Predictive data analitics is used for making predictions. Machine learning is closely related to this branch of analytics and shares almost all algorithms. Pattern Recognition It is an engineering field. It uses machine learning algorithms to extract the features for pattern analysis and pattern classification. One can view pattern recognition as a specific application of machine learning. ‘These relations are summarized in Figure 1.4. Data scence Data Data ‘mining analytics Machine learning Pattern Pe recognition rae ieee Figure 1.4: Relationship of Machine Learning with Other Major Fields 1.3.3 Machine Learning and Statistics Statistics isa branch of mathematics thathasa solid theoretical foundation regarding statistical leans Like machine learning (ML), it can lear from data, But the difference between statistics and ML is that statistical methods look for regularity in data called patterns, Initially, statistics sets a hypothesis a" performs experiments to verify and validate the hypothesis in order to find relationships among 42" 1\ re ee 6 © Machine Learning, am Big Data Data scionce concems about collection of data. Big data is a field of data science thay deals with data’s following characteristics: Volume: Huge amount of data is generated by big companies like Facebook, Twitt, YouTube. Variety: Data is available in variety of forms like images, videos, and in different formats . Velocity: It refers to the speed at which the data is generated and processed. Big data is used by many machine learning algorithms for applications such as language trans, lation and image recognition. Big data influences the growth of subjects like Deep learning, Deep learning is a branch of machine learning that deals with constructing models using neural networks en Data Mining Data mining’s original genesis is in the business. Like while mining the earth one gets into precious resources, it is often believed that unearthing of the data produces hidden infor. mation that otherwise would have cluded the attention of the management. Nowadays, many consider that data mining and machine learning are same. There is no difference between these fields except that data mining aims to extract the hidden patterns that are present in the data, whereas, machine learning aims to use it for prediction. Data Analytics Another branch of data science is data analytics. It aims to extract useful knowledge from crude data. There are different types of analytics. Predictive data analytics is used for making predictions. Machine learning is closely related to this branch of analytics and shares almost all algorithms. Pattern Recognition It is an engineering field. It uses machine learning algorithms to extract the features for pattern analysis and pattern classification. One can view pattern recognition as a specific application of machine learning. These relations are summarized in Figure 1.4. Data science Data Data mining analytics Machine learning Pattern recognition mane Re, Figure 1.4: Relationship of Machine Learning with Other Major Fields 1.3.3 Machine Learning and Statistics Statistics isa branch of mathematics that has a solid theoretical foundation regarding statistical learning Like machine learning (ML, it can learn from data, But the difference between statistics and ML is that statistical methods look for regularity in data called patterns. Initially, statistics sets a hypothesis a%4 performs experiments to verify and validate the hypothesis in order to find relationships among 42TTT Introduction to Machine Learning © 7 Statistics requires knowledge of the statistical procedures and the guidance of a good statistician. Itis mathematics intensive and models are often complicated equations and involve many assumptions. Statistical methods are developed in relation to the data being analysed. In addition, statistical methods are coherent and rigorous. It has strong theoretical foundations and interpretations that require a strong statistical knowledge. Machine learning, comparatively, has less assumptions and requires less statistical knowledge. But, it often requires interaction with various tools to automate the process of learning. Nevertheless, there is a school of thought that machine learning is just the latest version of ‘old Statistics’ and hence this relationship should be recognized. 1.4 TYPES OF MACHINE LEARNING What does the word ‘learn’ mean? Learning, like adaptation, occurs as the result of interaction of the program with its environment. It can be compared with the interaction between a teacher and a student. There are four types of machine learning as shown in Figure 1.5. Machine learning Se ee 2 nr 2 Supervised Unsupervised Semi-supervised Reinforcement learning learning learning learning | | [eae Cluster || Association || Dimension analysis mining || reduction [classification] Regression Figure 1.5: Types of Machine Learning Before discussing the types of learning, it is necessary to discuss about data, Labelled and Unlabelled Data Data is a raw fact. Normally, data is represented in the form of a table. Data also can be referred to as a data point, sample, or an example. Each row of the table represents a data point. Features are attributes or characteristics of an object. Normally, the columns of the table are attributes. Out of all attributes, one attribute is important and is called a label. Label is the feature that we aim to predict. Thus, there are two types of data — labelled and unlabelled. Labelled Data To illustrate labelled data, let us take one example dataset called Iris flower dataset or Fisher's Iris dataset, The dataset has 50 samples of Iris ~ with four attributes, length and width of sepals and petals. The target variable is called class. There are three classes — Iris setosa, Iris, virginica, and Iris versicolor. The partial data of Iris dataset is shown in Table 1.1.a | Be Blaine bearing Table 1.1: Iris Flower Dataset (Sic eu) eee oh eel 1. 55 42 14 02 Setosa 2 7 32. 47 14 Versicolor 3. 73 29 63 18 Virginica A dataset need not be always numbers. It can be images or video frames. Deep neural Networks can handle images with labels. In the following Figure 1.6, the deep neural network takes images, of dogs and cats with labels for classification. Input Label dog (b) Figure 1.6: (a) Labelled Dataset (b) Unlabelled Dataset In unlabelled data, there are no labels in the dataset, 1.4.1 Supervised Learning Supervised algorithms use labelled dataset. As the name suggests, there is a supervisor or teacher component in supervised learning. A supervisor provides labelled data so that the model is constructed and generates test data. In supervised learning algorithms, learning takes place in two stages. In layman terms, during the first stage, the teacher communicates the information to the student that the student is supposed '0 master. The student receives the information and understands it, During this stage, the teacher has n° knowledge of whether the information is grasped by the student. This leads to the second stage of learning, The teacher then asks the student a set of question’ to find out how much information has been grasped by the student, Based on these questions| 8 + Machine Learning —————<—<$<—$$$_$_$ a Table 1, Iris Flower Dataset Setosa 2 7 32. 14 Versicolor 3. 73 29 18 Virginica A dataset need not be always numbers. It can be images or video frames. Deep neural network, can handle images with labels, In the following Figure 1.6, the deep neural network takes images o¢ dogs and cats with labels for classification. Input Label dog cat Figure 1.6: (a) Labelled Dataset (b) Unlabelled Dataset In unlabelled data, there are no labels in the dataset, 1.4.1 Supervised Learning Supervised algorithms use labelled dataset. As the name suggests, there is a supervisor-or teacher component in supervised learning. A supervisor provides labelled data so that the model is constructed and generates test data, In supervised learning algorithms, earning takes place in two stages. In Jayman terms, during the first stage, the teacher communicates the information to the student that the student is supposed !° taster. The student receives the information and understands it. During this stage, the teacher has 00 knowledge of whether the information is grasped by the student, This leads to the second stage of learning. The teacher then asks the student a set of questio® to find out how much information has been grasped by the student. Based on these questions,‘oduction to Machine L the student is tested, and the teacher informs the student about his assessment. This kind of learning, is typically called supervised learning. pervised learning has two methods: 1, Classification 2. Regression Classification Classification is a supervised learning method. The input attributes of the classification algorithms are called independent variables, The target attribute is called label or dependent variable, The relationship between the input and target variable is represented in the form of a structure which is called a classification model, So, the focus of classification is to predict the ‘label’ that ina discrete form (a value from the set of finite values). An example is shown in Figure 1.7 where a classification algorithm takes a set of labelled data images such as dogs and cals to construct a model that can later be used to classify an unknown test image data. Ge) G6) Ze Labelled S| New test data data Classification Classification —} algorithm = [| model Label is Cat Figure 1.7: An Example Classification System In classification, learning takes place in two stages. During the first stage, called training stage, the learning algorithm takes a labelled dataset and starts learning. After the training set, samples are processed and the model is generated. In the second stage, the constructed model is tested with test or unknown sample and assigned a label. This is the classification process. This is illustrated in the above Figure 1.7. Initially, the classification learning algorithm learns with the collection of labelled data and constructs the model. Then, a test case is selected, and the model assigns a label. Similarly, in the case of Iris dataset, if the test is given as (6.3, 2.9, 5.6, 1.8, 2), the classification will generate the label for this. This is called classification. One of the examples of classification is — Image recognition, which includes classification of diseases like cancer, classification of plants, etc. The classification models can be categorized based on the implementation technology like decision trees, probabilistic methods, distance measures, and softcomputing methods. Classification models can also be classified as generative models and discriminative models. Generative models deal with the process of data generation and its distribution. Probabilistic models are examples ofag a Mil MN nr gm zeneration of data. Inste, Benerative models, Discriminative models do not care about the generation of data. Instead, thy simply concentrate on classi fying the given data, Some of the key algorithms of classification are: * Decision Tree * Random Forest * Support Vector Machines * Naive Bayes * Attificial Neural Network and Deep Learning networks like CNN Regression Models Regression models, unlike classification algorithms, predict continuous variables like price In other wonds, it is a number. A fitted regression model is shown in Figure 1.8 for a dataset that represent weeks input x and product sales y. a 1s+— y-axis - Product sales data (y) 1 2 3 4 5 axis - Week data (x) — Regression line (y = 0.66X +054) Figure 1.8: A Regression Model of the Form y= ax +b The regression model takes input x and generates a model in the form of a fitted line of the form y = fix). Here, x is the independent variable that may be one ot more attributes and y is the dependent variable. In Figure 1.8, linear regression takes the training set and tries to fit it with a line ~ product sales = 0.66 x Week + 0.54. Here, 0.66 and 0.54 are all regression coefficients that are learnt from data, The advantage of this model is that prediction for product sales (y) can be made for unknown week data (x). For example, the prediction for unknown eighth week can be made bY substituting x as 8 in that regression formula to get y. One of the most important regression next section. (Lioth regression and classification models are supervised algorithms. Both have a supervisor and the Concepts of training and testing are applicable to both. What is the difference between classification and regression models? The main difference is that regression models predict continuous variables such as product price, while classification concentrates on assigning labels such as clase - algorithms is linear regression that is explained in theoo Introduction to Machine Learning + 11 1.4.2 Unsupervised Learning The second kind of learning is by self-instruction. As the name suggests, there are no supervisor or teacher components. In the absence of a supervisor or teacher, self-instruction is the most common Kind of learning process. This process of self-instruction is based on the concept of trial and error. Here, the program is supplied with objects, but no labels are defined. The algorithm itself observes the examples and recognizes patterns based on the principles of grouping. Grouping is done in ways that similar objects form the same group. Cluster analysis and Dimensional reduction algorithms are examples of unsupervised algorithms. Cluster Analysis Cluster analysis is an example of unsupervised learning. It aims to group objects into disjoint clusters or groups. Cluster analysis clusters objects based on its attributes. All the data objects of the partitions are similar in some aspect and vary from the data objects in the other partitions significantly. Some of the examples of clustering processes are — segmentation of a region of interest in an image, detection of abnormal growth in a medical image, and determining clusters of signatures in a gene database. An example of clustering scheme is shown in Figure 1.9 where the clustering algorithm takes a set of dogs and cats images and groups it as two clusters-dogs and cats. It can be observed that the samples belonging to a cluster are similar and samples are different radically across clusters. = Py wf ch eee xf Lys rs cue) Ww mV Figure 1.9: An Example Clustering Scheme Some of the key clustering algorithms are: + kemeans algorithm * Hierarchical algorithms Bi )12 | * Machine Learning Dimensionatity Re Dimensionali i f fre onallity reduction algorithms are examples of unsupervised algorithms, [¢ takes dimension g : sion data as input a s the or dimension by taking advanta OF the data, hig sia get ane outputs the data in lower dimension by taking BE Of the aki Bhey ask of reducing the dataset with few features without losing the Beneralty “ang The differences by | Table 1.2, tveen supervised and unsupervised learning are listed in the following Table 1,2: Differences between Supervised and Unsupervised Learning (teehee) Thero is a supervisor component | No supervisor component jE supervisor component _| 1 Uses Labelled data Uses Unlabelled data peda Assigns categories or labels Performs grouping process such that similar g will bein one cluster jects ——_| 1.4.3 Semi-supervised Learning There are circumstances where the dataset has a huge collection of unlabelled data and son labelled data. Labelling isa costly process and difficult to perform by the humans, Semi-superigg algorithms use unlabelled data by assigning a pseudo-label. Then, the labelled and Pseudo-labelleg dataset can be combined j 1.4.4 Reinforcement Learning 5 Reinforcement learning mimics human ‘beings. Like human beings use ears and eyes to perceive the 1 world and take actions, reinforcement leaming allows the agent to interact with the environment y to get rewards. The agent can be human, animal, robot, or any independent program. The rewards enable the agent to gain experience, The agent aims to maximize the reward. The reward can be positive or negative (Punishment). When the rewards are more, the behavior gets reinforced and learning becomes possible. Consider the following example of a Grid game as shown in Figure 1.10. Block Y), Goal Danger Figure 1.10: A Grid game324 zw. Mate ening Dimensionality Reduction ss ised algorithms. It tak Dimensionatity reduction algorithms are examples of unsupers6 “Sa dimension data as input rd utp the data in Tower dimension by taking advantage of the Va ‘Bhgg ‘ ing th ian, Of the data. It is a task of reducing the dataset with few features — © generality, Ne Fearning.are listed in the following The differences between supervised and unsupervised Table 1.2. Table 1,2: Differences between Supervised and Unsupervised Learning SS nuanclemane) i qi 1. | Thereisa supervisor component _ | No supervisor componen ze Uses Labelled data Uses Unlabeled data = Assigns categories or labels Performs grouping process such that similar objects will be in one cluster 1.4.3 Semi-supervised Learning ige collection of unlabelled data and som toperform by the humans. Semi-superviseg label. Then, the labelled and psetido-labelleg There are circumstances where the dataset has a hy labelled data. Labelling is a costly process and difficult algorithms use unlabelled data by assigning a pseudo dataset can be combined. 1.4.4 Reinforcement Learning Reinforcement learning mimics human beings. Like human beings use ears and eyes to perceive the world and take actions, reinforcement learning allows the agent to interact with the environment to get rewards. The agent can be human, animal, robot, or any independent program. The rewards enable the agent to gain experience. The agent aims to maximize the reward. ‘The reward can be positive or negative (Punishment). When the rewards are more, the behavior gets reinforced and learning becomes possible. Consider the following example of a Grid game as shown in Figure 1.10. Block Goal Danger Figure 1.10: A Grid game—— introtuction to Machine Learning + 13 In this grid game, the gray tile indicates the danger, black is a block, and the tile with diag lines is the goal. The aim is to start, say from bottom-left grid, using the actions left, right, top and pottom to reach the goal state. To solve this sort of problem, there is no data. The agent interacts with the environment to get experience, In the above case, the agent tries to create a model by simulating many paths and finding rewarding paths. This experience helps in constructing a model there is no supervisor or gonal compared to supervised learning, It can be said in summary ions need to be taken to reach the final decision. Therefore, labelled dataset. Many sequential decis einforcement algorithms are reward-based, goal-oriented algorithms Scan for information on "Important Machine Learning Algorithms’ 1.5 CHALLENGES OF MACHINE LEARNING What are the challenges of machine learning? Let us discuss about them now. Problems that can be Dealt with Machine Learning erforming tasks like computation. For example, while calculating aman may blink but computers can display the result in GO, and even beat professional players of that game. Computers are better than humans in the square root of large numbers, an average hu seconds. Computers can play games like ches However, humans are better than computers in many aspects like recognition. But, deep ‘as well. Machines can recognize human better as machine learning systems still pends on the learning systems challenge human beings in this aspect faces ina second. Still, there are tasks where humans are require quality data for model construction. The quality of a learning system de quality of data. This is a challenge. Some of the challenges are listed below: 1. Problems—Machine learning can deal with the’ well-posed’ problems where specifications are complete and available. Computers cannot solve ‘ill-posed’ problems. Consider one simple example (shown in Table 1.3): Table 1.3: An Example 11 21 Bal 41 51 5 Cana model for this test data be multiplication? Thats, y=, x x,, Well! Itis true! But, this is. equally true that y may be y= x) *% OF y=2,%. So, there are three functions that fit the data. This means that the problem is ill-posed. To solve this problem, one needs more example to heck the model. Puzzles and games that do not have sufficient specification may become an ii-posed problem and scientific computation has many ill-posed problems,14 + Machine Learning it ie machine learning, Availability of a quay hhould be large and should not have day, 2. Huge data — This is a primary requirement of data is a challenge. A quality data means it s| problems such as missing data or incorrect data rn jig Data, 3. High computation power ~ With the availability of Big, oat requirement has also increased, Systems with Graphies a ae Processing Unit (TPU) are required to execute machine learn Be ia laceasetore learning tasks have become complex and hence time complex! ty has ised, and that can be solved only with high computing power. the computational resourey ng Unit (GPU) oF even Tesoy algorithms. Also, m, aching 4, Complexity of the algorithms ~ The selection of algorithms, describing the algorithns, application of algorithms to solve machine learning task, and omen algorthng have become necessary for machine learning or data scientists now. oi ms have become a big topic of discussion and it isa challenge for machine learning professionals to design, select, and evaluate optimal algorithms, 5. Bias/Variance ~ Variance is the error of the model. This leads to a problem called biasy variance tradeoff. A model that fits the training data correctly but fails for test dala, in general lacks generalization, is called overfitting. The reverse problem is called underfitting where the model fails for training data but has good generalization. Overfitting and underfitting are great challenges for machine learning algorithms. 1.6 MACHINE LEARNING PROCESS ‘The emerging process model for the data mining solutions for business organizations is CRISP-DM. Since machine learning is like data mining, except for the aim, this process can be used for machine learning. CRISP-DM stands for Cross Industry Standard Process - Data Mining. This process involves six steps. The steps are listed below in Figure 1.11. Understand the [*L_] Understand the business [>| data | ames Data Preprocessing LF £ Modelling | Model evaluation a Model deployment, Figure 1.11: A Machine LearningiData Mining ProcessIntroduction to Machine Learning + 15 + Understanding the business ~ This step involves understanding the objectives and requirements of the business organization. Generally, a single data mining algorithm is enough for giving, the solution. This step also involves the formulation of the problem statement for the data mining, process Understanding the data ~ It involves the steps like data collection, study of the charac: ae of the data, formulation of hypothesis, and matching of patterns to the selected hypothesis. 2. 3, Preparation of data ~ This step involves producing the final dataset by cleaning the raw data and preparation of data for the data mining process. The missing values may cause problems during both training and testing phases. Missing data forces classifiers to produce inaccurate results. This is a perennial problem for the classification models. Hence, suitable strategies should be adopted to handle the missing data 4, Modelling ~ This step plays a role in the application of data mining algorithm for the data fo obtain a model or pattern. evaluate ~ This step involves the evaluation of the data mining results using statistical analysis and visualization methods. The performance of the classifier is determined by evaluating the accuracy of the classifier. The process of classification is a fuzzy issue. For example, classification of emails requires extensive domain knowledge and requires domain experts. Hence, performance of the classifier is very crucial. 6. Deployment ~ This step involves the deployment of results of the data mining algorithm to improve the existing process or for a new situation. 1.7 MACHINE LEARNING APPLICATIONS Machine Learning technologies are used widely now in different domains. Machine learning app! cations are everywhere! One encounters many machine learning applications in the day-to-day life. Some applications are listed below: 1. Sentiment analysis - This is an application of natural language processing (NLP) where the words of documents are converted to sentiments like happy, sad, and angry which are captured by emoticons effectively. For movie reviews or product reviews, five stars or one star are automatically attached using sentiment analysis programs. 2. Recommendation systems - These are systems that make personalized purchases possible. For example, Amazon recommends users to find related books or books bought by people who have the same taste like you, and Netflix suggests shows or related movies of your taste. The recommendation systems are based on machine learning. Voice assistants ~ Products like Amazon Alexa, Microsoft Cortana, Apple Siri, and Google Assistant are all examples of voice assistants. They take speech commands and perform tasks. These chatbots are the result of machine learning technologies. Technologies like Google Maps and those used by Uber are all examples of machine learning, which offer to locate and navigate shortest paths to reduce time. ‘The machine learning applications are enormous. The following Table 1.4 summarizes some of the machine learning applications.eerie) Table a 5 Applications Survey : Applications - iness firm : ; nkruptey of a Busines rr Business Predicting the bankrupt jters and detecting credit card frauds i cdiction of bank loan defats - 2. | Banking Prediction of a entfenion, I08e classification, ag 3. Image Processing | Image search engines, obj (ieee i ic images generating synthetic imag: 4 | Audio/Voice Chatbots like Alexa, Mierosoft Cortans Developing chatbots op 3 customer support, speech (0 text, and text to voice So tatccommaniy | Trend analysis and identification of bogus calls, fraudulent calls ang cation its callers, churn analysis 6. Marketing Retail sales analysis, market basket analysis, Profit Performance analysis, market segmentation analySis, and study of travel patterns of customers for marketing tours 7__| Games Game programs for Chess, GO, and Atari video games ion, and sentiment analysis & | Natural Language | Google Translate, Text summarizati : Translation 9. | Web Analysis and | Identification of access pattems, detection of e-mail spams, viruses, ii Services personalized web services, search engines like Google, detection of promotion of user websites, and finding loyalty of users after web page layout modification 10. | Medicine Prediction of diseases, given disease symptoms as cancer or diabetes. Prediction of effectiveness of the treatment using patient history and Chatbots to interact with patients like IBM Watson uses machine learning technologies. Ti. | Multimedia and | Face recognition/identification, biometric projects like identification Security of a person from a large image or video database, and applications involving multimedia retrieval 12. | Scientific Domain | Discovery of new galaxies, identification of groups of houses based ‘on house type/geographical location, identification of earthquake epicenters, and identification of similar land use 1. Machine learning can enable top management of an organization to extract the knowledge from the data stored in various archives to facilitate decision making. Machine learning is an important subbranch of Artificial Intelligence (Al). ‘A model is an explicit description of patterns within the data. ‘A model can be a formula, procedure or representation that can generate data decisions. ions. see Humans predict by remembering the past, then formul late the strat ach same manne, the computers can pred by following the proces tegy and make a prediction. In the 6. Machine learning is an important branch of AI. Al is a much broader subject, The aim of Ali © . The aim develop intelligent agents. An agent can be a robot, humans, or other autonot mous systems.&SaoMoH|_§€ i Introduction to Machine Learning + 17 7. Deep learning is a branch of machife leaning, The difference between machine learning and deep learning is that models are constructed using neural network technology in deep learning. Neural networks are models constructed based on the human neuron models. Data science deals with gathering of data for analysis. It is a broad field that includes other fields. 9. Data analytics aims to extract useful knowledge from crude data. There are many types of analytics. Predictive data analytics is an area that is dedicated for making predictions. Machine learning is closely related to this branch of analytics and shares almost all algorithms. 10. One can say thus there are two types of data ~ labelled data and unlabelled data, The data with a label is called labelled data and those without a label are called unlabelled data, 11. Supervised algorithms use labelled dataset. As the name suggests, there is a supervisor or teacher component in supervised learning. A supervisor provides the labelled data so that the model is constructed and gives test data for checking the model. 12. Classification is a supervised learning method. The input attributes of the classification algorithms are called independent variables. The target attribute is called label or dependent variable. The relationship between the input and target variables is represented in the form of a structure which is called a classification model. 1S. Cluster analysis is an example of unsupervised learning, It aims to assemble objects into disjoint clusters or groups. . 15. Semi-supervised algorithms assign a pseudo-label for unlabelled data. Reinforcement learning allows the agent to interact with the enviconment to get rewards. The agent can be human, animal, robot, or any independent program. The rewards enable the agent to gain ‘experience. 16. The emerging process model for the data mining solutions for business organizations is CRISP-DM. This model stands for Cross Industry Standard Process - Data Mining. 17. Machine Learning technologies are used widely now in different domains. “Key Terms: Machine Learning — A branch of AI that concerns about machines to I being explicitly programmed. + Data~A raw fact. learn automatically without ‘+ Model ~ An explicit description of pattems in a data, Experience — A collection of knowledge and heuristics in humans and hist. torical training data in case of machines. Predictive Modelling ~ A technique of developing models and making a prediction of unseen data. Deep Leaming ~ A branch of machine learning that deals with constructing models using neural networks. * Data Science ~A field of study that encompasses capt turing of data to its analysis covering all stages of data management. * Data Analytics ~ A field of study that deals with analysis of data, * Big Data— A study of data that has characteristics of volume, variety, and velocity. Pattern Recognition — A field of study that analyses a pattern using machine learning algorithms.18 Machine bearing $$ @a a a@a a a Statistics — A branch of mathematics that deals with learning from data using statistical methods. Hypothesis - An initial assumption of an experiment. Leaming - Adapting to the environment that happens because of interaction of an agent with the environment, Label ~ A target attribute. Labelled Data — A data that is associated with a label Unlabeled Data ~ A data without labels. Supervised Learning ~ A type of machine learning that uses labelled data and learns with the help of a supervisor oF teacher component Classification Program — A supervisory leaning method that takes an unknown input and assigns a label for it. In simple words, finds the category of class of the input attributes. Regression Analysis ~ A supervisory method that predicts the continuous variables based on the input variables. Unsupervised Leaming - A type of machine leaning that uses unlabelled data and groups the attributes to clusters using a trial and error approach. Cluster Analysis - A type of unsupervised approach that groups the objects based on attributes so that similar objects or data points form a cluster. Semi-supervised Learning ~ A type of machine learning that uses limited labelled and large unlabelled data. It first labels unlabelled data using labelled data and combines it for learning Purposes. Reinforcement Learning ~ A type of machine learning that uses agents and environment interaction for creating labelled data for learning. Well-posed Problem - A problem that has well-defined specifications. Otherwise, the problem is called ill-posed. Bias/Variance - The inability of the machine leaning algorithm to predict correctly due to lack of generalization is called bias. Variance is the error of the model for training data. This leads to problems called overfitting and underfitting. Model Deployment ~ A method of deploying machine learning algorithms to improve the existing business processes for a new situation. SO eon eae Why is machine learning needed for business organizations? List out the factors that drive the popularity of machine learning. What is a model? Distinguish between the terms: Data, Information, Knowledge, and Intelligence. How is machine learning linked to AI, Data Science, and Statistics? List out the types of machine learning. Nowe pepo List out the differences between a model and pattern, Patterns are local and model is global for entire dataset — Justify. 8, Are classification and clustering are same or different? Justify,18 Machine Learning —___ Statistics ~ A branch of mathematics that deals with learning from data using statistical methods, Hypothesis ~ An initial assumption of an experiment. Learning ~ Adapting to the environment that happens because of interaction of an agent with the environment. Label ~ A target attribute, Labelled Data - A data that is associated with a label, Unlabelled Data ~ A data without labels, Supervised Learning - A type of machine learning that uses labelled data and learns with the help of a supervisor or teacher component. Classification Program — A supervisory leaning method that takes an unknown input and assigns alabel for it. In simple words, finds the category of class of the input attributes, Regression Analysis - A supervisory method that predicts the continuous variables based on the input variables. Unsupervised Leaming - A type of machine leaning that uses unlabelled data and groups the attributes to clusters using a trial and error approach. Cluster Analysis — A type of unsupervised approach that Broups the objects based on attributes So that similar objects or data points form a cluster. ‘ype of machine learning that uses limited labelled and large 8 labelled data and combines it for learning purposes. Reinforcement Learning ~ A type of machi for creating labelled data for learning. Well-posed Problem - A called ill-posed. ine learning that uses agents and environment interaction Problem that has well-defined specifications. Otherwise, the problem is Bias/Variance ~ The inability of the machine learning algorithm to Predict correctly due to lack of generalization is called bias. Variance is th error of the model for training data. This leads to Problems called overfiting and underfitting. Model Deployment — A method of de business processes for a new situation, Ploying machine learning algorithms to improve the existing Why is machine leaning needed for business organizations? Uist out the factors that drive the popularity of machine leaming, . What is a model? + Distinguish between the terms: Data, Information, Knowledge, and Intelligence, - How is machine learning linked to AI, Data Science, and Statistics? List out the types of machine learning, Nae wre . List out the differences between a model and Pattern. Patterns are local and model is global for entire dataset - Justify. + Are classification and clustering are same or different? Justity.oo Introduction to Machine Learning » 19 List out the differences between label 10. Point out the differences between supervised and unsupervised learning. 11. What are the differences between classification and regression? 2 12, Whatis a semi-supervised learning? 13. List out the differences between reinforced learning and supervised learning. 14. List out important classification and clustering algorithms. 15. List out at least five major applications of machine learning. Long Questions 1. Explain in detail the machine learning process model. 2. List out and briefly explain the classification algorithms. 3. List out and briefly explain the unsupervised algorithms. Numerical’Problems and Activities 1. Let us assume a regression algorithm generates a model y = 0.54 + 0.66 x for data pertaining to week sales data of a product. Here, x is the week and y is the product sales, Find the prediction for the 5% and 8" week, 2. Give two examples of patterns and models. 3. Survey and find out atleast five latest applications of machine learning. 4, Survey and list out atleast five products that use machine learning.