Saurabh Mishra

Engineered robust mushroom classification with 81.6% testing accuracy through genetic algorithm-driven neural network evolution across 100 generations using selection, crossover, and mutation. Boosted model performance on complex categorical data through a comprehensive preprocessing pipeline implementing 3 key techniques: label encoding, one-hot encoding, and scaling.

Developed a robust Sudoku puzzle solver using backtracking algorithm, which exhaustively searches for a valid completion and guarantees a solution if the puzzle is valid. Designed a modular code structure allowing for easy integration with different user interfaces and optimized the code for readability and maintainability.

Leveraged the Stanford Sentiment140 dataset, containing 1.6 million annotated tweets for model training and evaluation, applying extensive data cleaning and preprocessing techniques. Attained 78% accuracy in sentiment classification processing 1.6 million tweets through an efficient ML pipeline implementing Logistic Regression and Naive Bayes models.

Grade: 7.8 CGPA

Pursued a Master of Computer Applications degree, focusing on advanced topics in computer science. Achieved a CGPA of 7.8, demonstrating strong academic performance in the program.

Grade: 8.5 CGPA

Completed a Bachelor of Computer Applications, gaining foundational knowledge in computer applications. Maintained a high academic standing with a CGPA of 8.5.