AI_ML_Recovery_Week

This is 1-week AI/ML recovery plan — designed to rebuild confidence, reframe my learning, and help me feel the progress I'm making again.

Day 1

Tool's I have used so far in my ML journey :

Pandas - A library written for the Python programming language for data manipulation and analysis.
Numpy - A fundamental library for Python numerical computing.
Matplotlib - A open-source plotting library for Python.
Scikit-Learn - An open-source machine learning library for the Python programming language.
Tensorflow - An open-source framework for machine learning and artificial intelligence.
Streamlit - A open-source framework to rapidly build and share beautiful machine learning and data science web apps.

Concepts I have touched in Machine learning :

Regression - A technique used to capture the relationships between independent and dependent variables.
Classification - A supervised learning technique where models learn from labeled datasets to assign new data points to predefined categories or classes.
Decision Tree - A type of supervised machine learning algorithm used for classification and regression tasks.
Random Forest - A machine learning algorithm that combines multiple decision trees to create a more accurate and stable prediction.

Projects I have made so far :

DiagnoWise - Smart Disease Predicting App
DietForge – Diet planning web app using Flask
Suggestify – TV show recommender with Gemini API

Day 2

Goal: Reinforce fundamentals through tiny wins.

✅ Linear Regression

Implemented a simple linear regression model using Scikit-learn on a synthetic dataset. Visualized best-fit line and compared predictions with actual values.

✅ Decision Tree Classifier

Used the Iris dataset to train a Decision Tree model. Visualized predictions and explored accuracy on test data.

✅ KMeans Clustering

Applied KMeans to random 2D data. Visualized cluster centers and grouped data points.

Day 3 - Sentiment Analyzer 💬🔍

A simple Streamlit web app that classifies the sentiment of a given sentence as Positive, Negative, or Neutral using a Logistic Regression model trained on textual data.

🚀 Features

Real-time sentiment prediction
TF-IDF vectorization for text processing
Emoji-based feedback
Streamlit UI for easy interaction

🛠️ Tech Stack

Python, Scikit-learn
TF-IDF, Logistic Regression
Streamlit

🔍 Day 4 – Visualizing Models

Visualized decision boundary of logistic regression
Plotted top 10 most influential words in sentiment classification
Understood how certain features drive predictions

🚀 Day 5 – ML Pipeline & App Optimization

✅ Created a complete ML pipeline using sklearn.pipeline.Pipeline
✅ Combined preprocessing (TF-IDF) and model (Logistic Regression) in one file
✅ Refactored Streamlit app to use the pipeline directly
✅ Saved and loaded the pipeline using pickle for deployment
✅ Improved app performance by reducing redundancy

🔗 Check out the pipeline code

🔮 What’s Next

Day 6: Make the repo portfolio-worthy with deployment

“Start where you are. Use what you have. Do what you can.” – Arthur Ashe

AkibDa/AI_ML_Recovery_Week