In this repository, I have added some of my mini-projects in AI-ML. Each mini-project is in a separate folder containing the necessary data files and code files (jupyter notebooks). Folder names same as the project names.

Mini-Projects:

1. Stock Prediction:
   • This project uses XGBoost to predict stock prices.
   • The dataset contains intraday stock prices of MSFT (15 min intervals). Historical data of most recent 5000 timestamps.
   • I did EDA, feature engineering (lag features, moving averages) and simulation of a realistic trading strategy based on the model predictions.
2. Feature Matching, distance measurement:
   • This project uses OpenCV to match features between two images.
   • It uses ORB feature detector.
   • First I did feature matching. Then I manually defined a bounding box having an object of interest in one image and used the matched features to find the same object in the second image. Then I find the distance between camera and the object using the matched features (by calculating angle of view and known focal length of my camera).
3. Image Processing:
   • This folder contains my implementations of various image processing algorithms taken from the book "Computer Vision - Algorithms and Applications" (2nd Edition) by Richard Szeliski. 3_1 indicates the chapter-number_section-number.
   • The algorithms include:
     • 3_1: Point operations, Histogram equalization, Gamma correction, CLAHE implementation
     • 3_2: Tinkering with kernels, sharpening, blurring, edge detection, steering filter.
     • 3_3: Non-linear filtering: Median filter, bilateral filter. Dialation, erosion, opening, closing, majority filter.
     • 3_4: FFT and Wiener filtering (did not implement)
     • 3_5: Image pyramids, wavelets, Downsampling.