AI

Module 1: Introduction to Artificial Intelligence

Explain the concept of Artificial Intelligence (AI) and its significance in various fields.
Introduction of different types of AI, including Narrow AI and General AI.
Explore the practical applications of artificial intelligence (AI) in
various real-world scenarios and examine the consequential impact
on society.

Article: https://www.ibm.com/topics/artificial-intelligence
Book: "Artificial Intelligence: A Guide for Thinking Humans" by Melanie Mitchell

Module 2: Python Basics

Fundamentals of Python programming language.

Data types, variables, and basic operations.

Introduction control structures like loops and conditional statements.
Provide instruction on the fundamentals of functions and demonstrate their practical
implementation using Python libraries.

Book: "Automate the Boring Stuff with Python" by Al Sweigart

Module 3: Data Wrangling Techniques

Examine the significance of data wrangling within the data science workflow.
 Explain data cleaning, transformation, and handling missing values.
Introduce data aggregation and merging techniques.
Explore the process of data reshaping utilizing tools such as pandas.

Blog Post: The Art of Data Wrangling in 2024: Techniques and Trends.

Module 4: Introduction to Neural Networks

Explain the basic concept of artificial neural networks (ANN).

Introduction of neural network, including input, hidden, and output layers.Describe the

Explore the significance of activation functions and backpropagation in the training

process of neural networks.
Discuss the applications of neural networks in different fields.

Book: Ripley, Brian D. (1996) Pattern Recognition and Neural Networks, Cambridge

Module 5: Tensorflow & Keras - Artificial Neural Networks (ANN)

Introduce Tensorflow and Keras as popular deep learning frameworks.

Teach how to build and train basic artificial neural networks using Keras.

Clarify the concept of loss functions and optimizers within the realm of
deep learning
Explore the phenomenon of overfitting and methods to mitigate it.

Blog: Simulated Spotify Listening Experiences for Reinforcement Learning

with TensorFlow and TF-Agents.
Video Reference : Deep Learning with Python and Keras - Full Course by
freeCodeCamp.org
Module 6: Convolutional Neural Networks (CNN)

Introduce the concept of Convolutional Neural Networks (CNNs).

E

Teach how to build and train CNN using Keras with real-world datasets.
Discuss transfer learning and fine-tuning for CNN models.

Video Series: Convolutional Neural Networks

Tutorial: Convolutional Neural Networks for Visual Recognition
Book: "Deep Learning" by Ian Goodfellow, Yoshua Bengio, and Aaron Courville

Module 7: Recurrent Neural Networks (RNN)

Explain the concept of Recurrent Neural Networks (RNN).

Discuss the importance of sequential data processing in RNNs.
Teach how to build and train RNN models using Keras.
Introduce applications like language modeling and sentiment analysis.

Tutorial: Understanding LSTM Networks

Blog Post: The Unreasonable Effectiveness of Recurrent Neural Networks
Online Course: Sequence Models by Andrew Ng

Module 8: Natural Language Processing (NLP)

Introduce Natural Language Processing (NLP) and its applications.

Explain text preprocessing techniques, such as tokenization and stemming.
Teach how to build NLP models for tasks like sentiment analysis and text generation.
Discuss popular NLP libraries like NLTK and spaCy.
 Tutorial: https://www.ibm.com/topics/natural-language-processing
Book: "Speech and Language Processing" by Daniel Jurafsky and James H. Martin
Online Course: Get started with artificial intelligence Learning path by IBM

Module 9: IBM Cloud & Watson AI Services

Introduce IBM Cloud and its AI services like Watson.

Demonstrate how to set up and use AI services in the IBM Cloud platform.
Discuss the benefits and limitations of cloud-based AI services.
Guide students through practical examples and projects.

IBM Cloud: IBM Cloud

Watson AI Services: IBM Watson

Module 10: Build & Deploy ML Application

Teach how to build end-to-end machine learning applications.

Discuss model deployment techniques, including REST APIs and web services.
Cover best practices for scaling and monitoring ML applications.
Encourage students to develop their own ML applications.

Blog Post: Getting started with apps using IBM cloud

Tutorial: Deployment, IBM documentation
Book: Essentials of Application Development on IBM Cloud by IBM Redbooks

Teaching Material

Module 1: Introduction to Artificial Intelligence

Hour 1:
1.
Definition and scope of AI.
Distinction between Narrow AI and General AI.
Historical background and key milestones.
2.
Exploration of AI's impact in industries like healthcare, finance, and transportation.
Pros and cons of AI adoption in various domains.
Ethical considerations and responsible AI development.

Artificial Intelligence, IBM docs

A Brief History of Artificial Intelligence
[Book] AI Infrastructure Reference Architecture IBM Systems

Hour 2:

3.
Introduction to Machine Learning as a subset of AI.
Supervised, Unsupervised, and Reinforcement Learning.
Examples of ML algorithms and use cases.
4.
Ethical considerations in AI decision-making.
Bias and fairness in AI models.
Addressing challenges like data privacy and security.

[Book] "Machine Learning Yearning" by Andrew Ng

[Article] What is AI ethics?

Hour 3:

5.
Emerging trends in AI research and development.
AI-driven innovations and their potential impact.
Career opportunities and skillsets for AI professionals.
6. Discussion on AI portrayals in movies and popular culture.Addressing misconceptions
and myths about AI.
 [Article] "Artificial Intelligence Revolution”, IBM
[Book] "Superintelligence: Paths, Dangers, Strategies" by Nick Bostrom
[Article] Artificial Intelligence, IBM

Module 2: Python Basics

Hour 1:

1.
Overview of Python as a high-level programming language.
Setting up Python environment and using an IDE.
Writing your first Python program.
2.
Explanation of fundamental data types (int, float, str, bool).
Creating and manipulating variables.
Type conversions and basic input/output.

Online Course: Data Analysis with Python by IBM

Interactive Platform: Jupyter Notebook
Documentation: Beginner's guide to Python, IBM by Steven Frankel

Hour 2:

3.
Understanding conditional statements (if, else, elif).
Iteration with loops (for and while).
Writing and using functions.
4. Introduction to popular Python
libraries (e.g., NumPy, Pandas).
Installing and importing libraries.
Basic operations with NumPy arrays and Pandas DataFrames.
 Online Course: Data Analysis with Python by IBM
Interactive Platform: Jupyter Notebook
Documentation: Beginner's guide to Python, IBM by Steven Frankell

Hour 3:

5.
Reading from and writing to files in Python.
Common file formats (e.g., CSV, JSON) and their manipulation.
Error handling using try-except blocks.
6.
Code organization and style guidelines (PEP 8).
Using virtual environments for project isolation.
Tips for debugging and troubleshooting.

Online Course: Data Analysis with Python by IBM

Interactive Platform: Jupyter Notebook
Documentation: Beginner's guide to Python, IBM by Steven Frankell

Module 3: Data Wrangling Techniques

Hour 1:

1.
Understanding the data wrangling process in the data science workflow.
Importance of data cleaning and preparation for analysis.
and c.
2.
Techniques for handling missing data and outliers.
Dealing with duplicates and inconsistent data.
Data normalization and scaling.
 Book: Securely Leverage Open-Source Software with Python AI Toolkit for IBM z/OS
Article: Python and Data Analytics workshop doc by IBM
Online Course: IBM Data Science Professional Certificate

Hour 2:

3.
Aggregating data using pandas groupby function.
Performing operations on grouped data.
Creating pivot tables for data summarization.
4.
Combining datasets using pandas merge function.
Understanding different types of joins (inner, outer, left, right).
Handling common merging challenges.

Book: Securely Leverage Open-Source Software with Python AI Toolkit for IBM z/OS
Article: Python and Data Analytics workshop doc by IBM
Online Course: IBM Data Science Professional Certificate

Hour 3:

5.
Reshaping data using pandas melt and pivot functions.
Converting data between long and wide formats.
Handling multi-indexing and hierarchical data.
6.
Students work on a data wrangling project.
Applying data cleaning, transformation, and aggregation techniques.

Book: Securely Leverage Open-Source Software with Python AI Toolkit for IBM z/OS
Article: Python and Data Analytics workshop doc by IBM
Module 4: Introduction to Neural Networks

Hour 1:

1. **Introduction to Neural

Networks**

Understanding biological neurons and artificial neurons.

Basic structure and working of neural networks.
Activation functions and their role in the network.

2.
Implementing a basic neural network from scratch in Python.
Forward and backward propagation for training.
Gradient descent and optimization techniques.

[Article] Neural Networks and Deep Learning - Link

[Book] "Make Your Own Neural Network" by Tariq Rashid
[Video] Neural Networks Demystified - Link

Hour 2:

3.
Understanding deep neural networks and hidden layers.
Popular architectures (e.g., CNN, RNN) and their applications.
Transfer learning and pre-trained models.
4.
Overview of TensorFlow and Keras as deep learning frameworks.
Building neural networks using Keras.
Compiling and training a model.

[Course] Build a Neural Network with pytorch

Hour 3:
 5.
Techniques for preventing overfitting (e.g., dropout, L2 regularization).
Optimizing hyperparameters for better performance.
Evaluating model performance with validation and test sets.
6.
Working with image data in deep learning models.
Case study: Image classification using CNN.
Challenges and best practices.

[Course] CNN using pytorch , IBM Skillsbuild

A Comprehensive Guide to Hyperparameter Tuning, IBM docs

Module 5: Data Visualization with Python

Hour 1:

1.
Importance of data visualization in data analysis.
Types of data visualizations and their applications.
Principles of effective data visualization.
2.
Creating line plots, bar charts, and scatter plots with Matplotlib.
Customizing plot appearance (e.g., labels, colors, legends).
Saving plots to various file formats.

Online Course: "Introduction to Data Science" on edX by IBM.

Hour 2:

3.
Introduction to Seaborn for statistical data visualization.
Creating bar plots, box plots, and violin plots with Seaborn.
Heatmaps and cluster maps for visualizing correlations.
4.
Building interactive plots and dashboards with Plotly.
Customizing Plotly visualizations and adding interactivity.
Embedding Plotly plots in web applications.

[Tutorial] Seaborn Official Tutorial - Link Book:

Securely Leverage Open-Source Software with
Python AI Toolkit , IBM

Hour 3:

5.
Introduction to geospatial data and geographic information systems (GIS).
Plotting geospatial data on maps with libraries like Geopandas and Folium.
Choropleth maps for visualizing regional data.
6.
Tips for designing effective and informative visualizations.
Students work on a data visualization project.
Presenting and sharing data visualizations.

Article: "IBM Data Science - Best Practices" by Thomas

[Tutorial] Data Visualization Best Practices, IBM

Module 6: Introduction to Machine Learning

Hour 1:

1.
Understanding the core concepts of machine learning.
Types of machine learning: Supervised, Unsupervised, and Reinforcement
Learning.
Real-world examples of machine learning applications.
2.
Explanation of regression and its applications.
Linear regression and multiple linear regression.
Evaluating regression models and interpreting results.

Online Course: Machine Learning with python by IBM

Book : Machine Learning For Dummies®, IBM Limited Edition

Hour 2:

3.
Understanding classification problems and algorithms.
Logistic regression and k-Nearest Neighbors (k-NN) algorithm.
Decision trees and random forests.
4.
Introduction to clustering and its applications.
k-Means clustering and hierarchical clustering.
Evaluating clustering performance.

Online Course: Machine Learning with python by IBM

Book : Machine Learning For Dummies®, IBM Limited Edition

Hour 3:

5.
Motivation for dimensionality reduction techniques.
Principal Component Analysis (PCA) and Singular Value Decomposition (SVD).
 t-Distributed Stochastic Neighbor Embedding (t-SNE).
6.
Understanding the end-to-end machine learning workflow.
Model evaluation and hyperparameter tuning.
Deploying machine learning models to production.

Online Course: Machine Learning with python by IBM

Book : Machine Learning For Dummies®, IBM Limited Edition

Module 7: Natural Language Processing (NLP)

Hour 1:

1.
Overview of NLP and its applications.
Challenges in processing human language.
Key NLP

tasks: Tokenization, Part-of-Speech (POS) tagging, and Named Entity Recognition (NER).

2.
Cleaning and normalizing text data.
Removing stopwords and special characters.
Lemmatization and stemming.

Tutorial: https://www.ibm.com/topics/natural-language-processing

Hour 2:

3.
 Bag-of-Words (BoW) model and Term Frequency-Inverse Document Frequency
(TF-IDF).
Word embeddings and distributed representations (Word2Vec, GloVe).
Language models and pre-trained transformers.
4.
Understanding sentiment analysis and its applications.
Building sentiment analysis models using machine learning and deep learning.
Interpreting sentiment analysis results.

Book: "Speech and Language Processing" by Daniel Jurafsky and James H. Martin
Online Course: Get started with artificial intelligence Learning path by IBMMM

Hour 3:

5.
Introduction to topic modeling and Latent Dirichlet Allocation (LDA).
Implementing LDA for discovering topics in text data.
Evaluating topic models and selecting the optimal number of topics.
6.
Introduction to text generation techniques.
Markov chains and Recurrent Neural Networks (RNNs) for text generation.
Building a simple text generation model.

Book: "Speech and Language Processing" by Daniel Jurafsky and James H. Martin
Online Course: Get started with artificial intelligence Learning path by IBMMMMMM

Module 8: Image Processing and Computer Vision

Hour 1:

1.
 Understanding the basics of image processing.
Image representation and color spaces.
Applications of computer vision in various domains.
2.
Convolution operation and image filtering.
Applying filters for blurring, sharpening, and noise reduction.
Edge detection techniques (e.g., Sobel, Canny).

Computer vision documentation, IBM

YOLO V8 Docs

Hour 2:

3.
Understanding object detection and localization.
Haar cascades and Single Shot MultiBox Detector (SSD).
Applying pre-trained models for object detection.
4.
Introduction to facial recognition techniques.
Face detection and face recognition algorithms.
Building a simple facial recognition system.

[Article] Introduction to computer vision

[Article] Computer Vision, IBM Research

Hour 3:

5.
Understanding image segmentation and its applications.
Thresholding and region-based segmentation.
Advanced segmentation methods (e.g., Watershed, GrabCut).
6.
Students work on an image processing project.
Applying various computer vision techniques.
 [Course] Classification with PyTorch, IBM
[Tutorial] Fine tune watson NLP

Module 9: Deep Learning: Advanced Topics

Hour 1:

1.
Introduction to RNNs and their architecture.
Vanishing and exploding gradients in RNNs.
Applications of RNNs in natural language processing and time series analysis.
2.
Understanding LSTM networks and their improvements over RNNs.
Building LSTM models for sequential data processing.
Applications of LSTMs in sentiment analysis and language translation.

[Tutorial] Pytorch, Tensor ,datasets, IBM

[Tutorial] Hugging face with IBM

Hour 2:

3.
Introduction to GANs and their architecture.
Training GANs using adversarial learning.
Applications of GANs in image generation and style transfer.
4.
Understanding transfer learning and its benefits.
Fine-tuning pre-trained models for specific tasks.
Implementing transfer learning with TensorFlow/Keras.

[Tutorial] GANs: Introduction and Implementation

 [Tutorial] Projects using DGCAN, IBM skills Network

Hour 3:

5.
Understanding autoencoders and their architecture.
Latent space representation and anomaly detection.
Variational Autoencoders for generative modeling.
6.
Students work on a deep learning project.
Applying advanced deep learning techniques.

[Tutorial] Autoencoders in TensorFlow

[Tutorial] Variational Autoencoders Explained

Module 10: Build & Deploy ML Applications

Hour 1:

1.
Understanding the need for model deployment.
Options for deploying machine learning models.
Challenges and considerations in deployment.
2.
Introduction

to Flask as a web application framework.

Setting up a basic Flask application.

Integrating machine learning models into Flask apps.

[Tutorial] IBM Watson Documentation

 [Tutorial] Flask Official Documentation

Hour 2:

3.
Introduction to Docker and containerization.
Creating Docker containers for machine learning models.
Deploying Dockerized applications.
4.
Understanding FastAPI as a modern API framework.
Building RESTful APIs with FastAPI.
Integrating machine learning models with FastAPI.

[Tutorial] IBM Cloud Pak Data Documentation

[Tutorial] API, IBM docs

Hour 3:

5.
Overview of cloud service providers (AWS, Azure, Google Cloud Platform).
Deploying machine learning models on the cloud.
Scalability and cost considerations.
6.
Students work on a complete ML application project.
Building, deploying, and presenting the application.

[Tutorial] Deploying Machine Learning Models - IBM Cloud

[Tutorial] Deployment using Kubernetes- IBM Cloud