Kunjan Mhaske's Projects
It contains the reward function and model generated that used in battle-of-schools March 2020 at RIT
π Papers & tech blogs by companies sharing their work on data science & machine learning in production.
A curated list of articles that cover the software engineering best practices for building machine learning applications.
Bag of Words model using SIFT (Scale-Invariant Feature Transform β 128 dimension vector) from OpenCV library to get features of images and cluster them using KMeans from sklearn library to create vocabulary of words for each type of images.
A complete computer science study plan to become a software engineer.
Implemented and compared NaΓ―ve Bayes and Decision tree using Python to classify genuine and counterfeit banknote.
Data Analysis dashboard developed with Python Django and Flexmonster
Fashion-MNIST model to detect the types of clothes using 2D Convolutional Neural Networks and Linear Transformations from PyTorch-gpu library with Cross Entropy as a Loss function, ReLu activation function and Adam optimizer for training the model.
The Open Source Data Science Masters
Image Super Resolution model with BSD300 dataset using 4 layers of 2D Convolutional Neural Networks from PyTorch-gpu library with MSELoss as a Loss function, ReLu activation function and Adam optimizer that achieves the best PSNR of 24.41 bB.
Mastering OpenCV 4 with Python, published by Packt
Python assignments for the machine learning class by andrew ng on coursera with complete submission for grading capability and re-written instructions.
Applied YOLO model trained on COCO dataset to detect obstacles and Lane-Net model trained on tusimple.ai dataset for end-to-end lane detection. β’ Improved usability and response time by 50% using the combination and optimization of legacy codes of algorithms in assisted driving.
This repository is intended for practice purpose only
Python Data Science Handbook: full text in Jupyter Notebooks
Implemented and compared Random Forest, Decision Tree, KNN, SVM, and Logistic Regression outcomes with a confusion matrix. Concluded that Random Forest achieved the highest accuracy of 85% to predict the loan status for investors.
AWiFS Dataset image processing python scripts to detect water bodies and burnt areas from multi-spectral sensor images.
K12 Schools and University Search Web Applications (Contains APIs and Frontend code using Node JS, Sequelize ORM, React JS, and MySQL)
Learn how to design large-scale systems. Prep for the system design interview. Includes Anki flashcards.
π― Materials to help you rock your next coding interview