Introduction to Artificial Intelligence and Machine Learning with Project-Based Learning

Course Description:

This comprehensive course introduces learners to the core Machine Learning (ML) concepts of Artificial Intelligence (AI), Machine Learning (ML), Supervised Learning and Classification Model covering both theoretical and practical aspects. With a focus on hands-on learning, the course is designed to walk you through the essential concepts of performance metrics, hyperparameter fine-tuning of ML Classification model and how to apply these concepts to real-world projects.

Through detailed explanations and project-based modules, this course breaks down the complex concepts into manageable sections, offering clear explanations and examples. You will work through a project divided into multiple modules, with each program (or section) explained in a step-by-step manner. By the end of the course, you would have built two Machine Learning Classification Models Viz Backend based ML classification Model (using first three Python Jupyter Notebook based programs) and Frontend based ML Classification model (Using two Python programs with Streamlit library and Pickle file).

This course ideally suits the professional who want to become expert in Machine Learning domain. Also, it suits well to the college students who wants to do their final year projects in ML domain besides making them job enabled in ML domain.

This course assumption is that the learners are aware of basic Python programming language.

Course Outline:

1. What is Artificial Intelligence?
   • A clear introduction to AI, its scope, and its real-world applications.
2. What is Machine Learning?
   • Learn how ML fits into the broader field of AI and the types of learning processes.
3. What is Supervised Learning?
   • Understanding supervised learning and its use cases in building predictive models.
4. What is an ML Classification Model?
   • Introduction to classification models and how they are used to predict the data based on the learning of classifier algorithms.
5. What are Classification Model Metrics?
   • An explanation of metrics like accuracy, precision, recall, F1 score, and more to evaluate ML Classification model performance.
6. Fine-Tuning Classifiers’ Hyperparameters
   • Learn how to optimize machine learning models by adjusting tuneable hyperparameters of the Classifier algorithms to improve accuracy and performance.
7. Project Overview:
   • Part-1: Project Problem Statement, Datasets and Proposed Solution
     • Overview of the machine learning project (Bank Housing Loan application status prediction), Problem Statement, Input and Customer Datasets formats and Solution proposed for solving the problem.
   • Part-02: Project Design and Tools and Technologies planned to be used in the project
     • Project or Solution Backend and frontend design explanations. Also, the Tools and Technologies which are planned to be used in the project.
8. Module 01/Program 01: Training Data Loading, Identifying the ML Model and Preprocessing
   • Introduction to the first module of the project.
   • Section 01: Data Loading and verification
     • Read the training data and load it as Dataframe. Identify the ML model applicable like Supervised Learning, Classification subtype. Check for missing values and duplicates if any.
   • Section 02: Data preprocessing
     • Preprocess data by dropping the insignificant columns, converting the object datatype columns to respective numerical data values and segregate the Labelled data and labels for using with ML model.
   • Section 03: Build and evaluate the ML Classification model with and without data scaling
     • Build and evaluate the ML Model using Logistic Regression algorithm and CV Method without scaling

• Build and evaluate the ML Model using Logistic Regression algorithm and CV Method with standardized scaling
• Build and evaluate the ML Model using Logistic Regression algorithm and CV Method with normalized scaling

1. • Section 04: Identify the suitable ML Classification model combination and write out the pre-processed Training file
     • Build and evaluate the ML Model using Logistic Regression algorithm and CV Method with finetuning of hyperparameters but without scaling
     • Identify suitable ML model combination based the ML Model metrics
     • Write out the pre-processed training file
2. Module 02/Program 02: Identifying the most suitable Classifier algorithm for the given training dataset
   • Introduction to the second module of the project.
   • Section 01: Load the preprocessed training data and define the hyperparameters of main classifier algorithms
     • Read the Preprocessed training data and prepare the data for ML model (Labelled data and labels by slicing method).
     • Define the tuneable Hyperparameters for all the main classifier algorithms
   • Section 02: Identify the most suitable classifier algorithm for the given dataset based on the ML Classification model metrics
     • Execute each classifier algorithm with given hyperparameters combination with Cross validation using the preprocessed training data and record the metrics
     • Identify the most suitable classifier algorithm for the given dataset based on the metrics value
3. Module 03/Program 03: Predicting the Loan application approval status values for the given Customer data and writing out the Pickle file
   • Introduction to the third module of the project.
   • Section 01: Data preparation (Combining Training and Customer datasets) for the ML Model
     • Read the Training and Customer datasets and combine them with flags so that they can be segregated later
   • Section 02: Preprocessing of the combined data
     • Preprocess the combined data
     • Segregate the preprocessed combined data into training and customer data
     • Prepare the preprocessed training data for ML Model
   • Section 03: Predict the Loan_Status values, write out the Customer data file with predicted values and Pickle file
     • Build and evaluate the ML Classification model with finalized classifier algorithm
     • Predict the Loan_Status values for the given Customer data
     • Write out the Customer data file with predicted values and Pickle file
4. Module 04/Program 04: Loan application status prediction using interactive frontend (Single record at a time)
   • Load and open the Pickel file in read binary mode
   • Define and display the frontend screen using Streamlit library using in a python program
   • Collect the Bank Housing Loan Application data (one data at a time) through the displayed frontend
   • Predict the Loan application approval status using the data collected through frontend with the Pickle file
   • Display the Loan application predicted status in the frontend
5. Module 05/Program 05: Loan application status prediction using interactive frontend (using a customer file as an input through frontend)
   • Load and open the Pickel file in read binary mode
   • Define and display the frontend screen using Streamlit library using in a python program
   • Browse the Customer file through the solution frontend from the local folder
   • Read and display the customer data (top 10 records)
   • Preprocess the Customer data
   • Predict the Loan application approval status for all the records given in the customer file using the Pickle file
   • Download and write the Customer file with the Predicted Loan application status values into any local folder
6. Project Setup
   • Download and setup the project files in your system
   • Download and install Anaconda Navigator tool in your system
   • Set the Streamlit environment in your Anaconda Navigator Tool
7. Project Execution Demos
   • Project Environment Setup:
   • Start the Anaconda Navigator Tool in your system
   • Open the VS Code Tool from Anaconda Navigator
   • From VS Code open your Project Folder
   • Module 01/Program 01 Execution:
   • Execute the Module01/Program01 from the VS Code Tool editor
   • Verify its successful execution and creation of “Preprocessed Training file” in the “Data-files” folder
   • Module 02/Program 02 Execution:
   • Execute the Module02/Program02 from the VS Code Tool editor
   • Verify its successful execution and note the most suitable classifier algorithm identified in this program display
   • Module 03/Program 03 Execution:

• Execute the Module03/Program03 from the VS Code Tool editor
• Verify its successful execution
• Verify the successful creation of Customer data file with predicted Loan_Status values in the “Backend-Output-files” folder
• Verify the successful creation of Pickle file in the “Pickle-File” folder

1. • Module 04/Program 04 Execution:

• Execute the Module04/Program04 from the VS Code Tool Terminal
• Verify the display of the Frontend in your System
• Enter the Bank Housing Loan application data and click on “Submit” button
• Verify the successful display of the Loan application approval status prediction in the frontend

1. • Module 05/Program 05 Execution:

• Execute the Module05/Program05 from the VS Code Tool Terminal
• Verify the display of the Frontend in your System
• Browse the Customer data file from the frontend
• Verify the successful display of the Customer data file with predicted Loan_Status values
• Verify the successful creation of Customer data file with predicted Loan_Status values in the “Frontend-Output-files” folder

By the end of this course, you will have a solid foundation in AI and machine learning and hands-on experience in building and executing machine learning project. This course is ideal for beginners and professionals who are looking to gain practical experience with AI and ML Model. This course is very ideal for college students those who wanted to do creative final year project which in turn enable them for the Job in ML domain.