A comprehensive training program covering the fundamentals of Python and data science, from basic programming to advanced machine learning model deployment. This repository contains the projects, code, and resources from the training.
This training program was a collaborative effort between IINTM college and Grastech. The entire curriculum was expertly designed and delivered by the trainer, Ankit Mishra sir.
Trainer's GitHub: https://github.com/ankitmisk
The training was structured to provide an in-depth, hands-on learning experience. It progressed logically through key concepts and technologies, culminating in the successful deployment of machine learning models.
Basic Programming: Covered basic data types, conditional statements (if, else, elif), and loops (for, while).
Functional & Module Programming: Explored functional programming concepts and created custom modules. Included projects on a Paytm Sound Box, Fibonacci series, anagrams, palindromes, and pattern printing.
File Handling: Practiced working with binary and textual files, including a project that scraped images from Lorem Picsum.
Web Scraping: Developed skills in web scraping by extracting real-world data, such as COVID-19 and Rich List data from Wikipedia.
Object-Oriented Programming (OOPs): Mastered the four pillars of OOPs (Encapsulation, Polymorphism, Inheritance, Abstraction), including Dunder/Magic methods, class constructors, and different types of inheritance. Included projects like a Car Class and an ATM Class.
Database Connectivity: Established connections between Python and MySQL, performing CRUD (Create, Read, Update, Delete) operations. Completed a mini-project to scrape web data and dump it into a database.
NumPy: Explored the creation and manipulation of N-dimensional arrays. Used array methods to create and modify images, and learned about array indexing, slicing, broadcasting, and key methods like shape, size, and reshape.
Pandas: Focused on data analytics and EDA (Exploratory Data Analysis). Learned to create Series and DataFrames, and used important methods like head, tail, describe, isna, corr, and fillna. Conducted in-depth EDA projects on the Car Dekho and IPL Datasets, including masking and DataFrame filtering.
Introduction to ML: Gained a clear understanding of the differences between Data Science, AI, ML, and Gen AI. Explored the benefits and applications of classical machine learning models.
Supervised Learning:
Regression: Implemented models like Linear Regression, Lasso, and Ridge. Understood the mathematical intuition behind error metrics (MSE, MAE, RMSE) and the concept of Gradient Descent.
Project: Developed and deployed a California House Price Prediction model on a local host and a live Streamlit server, focusing on bias vs. variance trade-offs. The best model was saved using Pickle.
Classification: Implemented various models including Logistic Regression, Naive Bayes, Support Vector Machine, and Decision Trees. Evaluated models using Confusion Matrix and metrics like Accuracy, Precision, Recall, and F1 score.
Project: Built and deployed a Heart Disease Prediction model on both a local host and a live Streamlit server. Explored different model approaches and compared their performance to identify the best one.
Model Deployment: The training concluded with live deployment of both regression and classification projects on a Streamlit server, making the projects publicly available.