Scribd
Upload
43 views3 pages

AI Bootcamp Notes

The AI Bootcamp outlines a four-stage learning process: Inspire, Acquire, Experience, and Empower, focusing on AI ethics, statistical data, and practical applications in NLP, CV, and data science. It details the AI project cycle, which includes problem scoping, data acquisition, exploration, modeling, evaluation, and deployment, emphasizing the importance of data accuracy and ethical considerations. Additionally, it describes three families of machine learning models: supervised, unsupervised, and reinforcement learning.

Uploaded by

mansasingh2007
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
43 views3 pages

AI Bootcamp Notes

The AI Bootcamp outlines a four-stage learning process: Inspire, Acquire, Experience, and Empower, focusing on AI ethics, statistical data, and practical applications in NLP, CV, and data science. It details the AI project cycle, which includes problem scoping, data acquisition, exploration, modeling, evaluation, and deployment, emphasizing the importance of data accuracy and ethical considerations. Additionally, it describes three families of machine learning models: supervised, unsupervised, and reinforcement learning.

Uploaded by

mansasingh2007
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

AI Bootcamp:

4 stages of learning:
● Inspire( AI ethics, bias, etc.)
● Acquire( Statistical data, NPL, CV)
● Experience(Data collection, model training, code modification)
● Empower(create projects)

3 domains of AI: NLP, CV, Data science (on basis of input)

NLP: domain of AI that uses data in the form of text, audio and speech. For e.g.sematris
Computer vision: Input as images/videos or related to vision. For eg.quick draw.
Data science: In the form of numbers. e.g.rock paper scissor.

Artificial Intelligence: When a machine is able to mimic human intelligence. Like solving real
world problems, taking decisions, improving from experience.

How does a machine mimic human intelligence?


● Through information fed in the form of data
● Develops an algorithm
● And learns through experience

Statistical data:
Data in the form of numerical and alpha- numerical data(both numbers and alphabets)

Applications:
score prediction( Stock markets)
Weather forecasting
Targeted Advertising

Computer Vision
Data in form of images and videos

Applications:
Face Lock
Object Classification
Filters in apps like snapchat

NLP:
Data in form of text and speech

Applications:
Chatbot
Google search
autocorrect

AI application around us:


Siri, Cortana, Lenskart, Swiggy, Youtube ,Flipkart, Alexa

SDG: 17 goals by UN in 2015

AI Ethics: Set of rules/ principles a machine follows

Data Privacy: Includes the permissions asked by websites...


AI bias: Machine becomes biased based on the data we train it with. If the data is biased,
output is adulterated.

AI access: Something you can reach.


Accessibility to AI tool sets so that you can learn from experience

If a machine is trained by data, when will data be inaccurate?


If we are cautious about data then we will be able to be accurate.

4/5/21
AI project cycle: A cycle/ process
Various stages:

1. Problem Scoping: Stage where you are identifying the problem

4Ws Problem Canvas: Who, what, where and why


Who: who is having the problem/ who are the stakeholders?
(any person/ community who is facing the problem)

What: what is the nature of the problem?


What is the problem?
1. rise in cabbage prices
2. unpredictability of cabbage prices

How do we know it is a problem?


news articles

Where? Where does the problem arise? What is the context?


during kimchi making seasons
during cabbage harvest seasons

Why? Why do you believe it is a problem worth solving?


What would be of key value to stakeholders?
to buy cabbage at low prices
How would it improve their situation?
Allow them to make more kimchi and less household expenses

We would put it all in the problem statement template.

2. Data Acquisition: Collecting information


we collect data on factors that are responsible for problem

3. Data exploration: Explore more on the collected Data/ organizing the data
To ensure authenticity we go through the government. websites and find relevant data.Explore
trends and patterns. Focuses on analyzing the data.

Now going to split the analyzed data into 2 parts: training data and testing data(usually 70-30)

4. Modeling: implementing the algorithm on the data to train/refine it.


5. Evaluation: testing the model
6. Deployment: implementing the solution in a real-life scenario.

Stage 1 to 5 is called development.


The AI project cycle basically has 2 stages: development and deployment.

Machine learning has 3 families of models:


● Supervised: Determine relationships through training. Learn through observation.
Learn through multiple examples. Trained using vast amounts of data.
Data is said to be labeled when all data is categorized.

● Unsupervised: Discover new patterns. AI models learn through transversing through


large amounts of data and classifying it. We feed unlabelled data and undergo training.
Useful when labeled data isn't available.Catches patterns that humans can't catch.

If you have clean labeled data: supervised learning and if it's difficult to sort data: unsupervised
learning.

● Reinforcement: Learn by rewarding actions(work in confined environment like


games)

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy