Suraj Varma's repositories
Protein-sequence-analysis
This notebook does the analysis of a protein sequence dataset in FASTA format. It addresses challenges in data import, alignment, embedding, and classification using bioinformatics tools and machine learning techniques. The solution provides a systematic approach to extract insights from structured data, crucial for bioinformatics research.
Bakery-Sales-Analysis-Dashboard-and-Prediction
Explore bakery sales data and AI predictions with Python. Analyze trends, visualize insights, and model product sales using pandas, matplotlib, and scikit-learn in this comprehensive GitHub repository.
Ball-Trajectory-Kalman-Filter
Simulates ball trajectory with gravity and bounce effects, applying a Kalman Filter for position estimation from noisy observations. Visualizes true trajectory, noisy data, and filter estimates with performance metrics.
Ball-Trajectory-Particle-Filter
Simulate and estimate the trajectories of two balls using particle filters. Includes noisy observations, particle filtering, error calculations, and visualizations. Requires Python, `numpy`, and `matplotlib`.
Bert-emabadding-generation
Clustering text data using BERT embeddings with AG News dataset from Hugging Face. Includes scripts for generating, preprocessing, and saving embeddings for advanced clustering and analysis tasks.
Burger-sales-data-and-report-generator
Automated Burgerking sales data generator in Python. Visualizes sales, ingredient consumption, and KPIs with interactive Plotly charts. Simulates one-year dataset for analysis.
Course-management-sysyem.
A C++ program for university course management, featuring `Course`, `Lecturer`, `Student`, and base `Person` classes. It supports course registration within 3-10 student limits, caters to both internal and external students, and includes a user interface menu. Focuses on object-oriented design, without needing data persistence.
Cuisine-Prediction
An NLP task to predict Cuisine from recipe data, visualize cuisine distribution, train MNB, XGBoost, CNN, and Random Forest models for cuisine prediction, and evaluate performance with accuracy, confusion matrix, and classification report.
Dynamic-Sales-and-Inventory-Dashboard-of-Burgerking
A fully dynamic dashboard visualizing synthetically generated data, providing real-time sales and ingredient consumption insights with interactive elements, peak hours analysis, and customizable views for efficient inventory and sales management.
Game-score-Management-System
Developed a C++ program for managing combat games, transitioning from paper to digital. Features include player data entry, game management, dynamic point allocation, and winner announcement. Utilizes operator overloading for data input/output and comparisons. Designed for up to 6 players, with in-memory data storage during runtime.
Gujarati-Song-Lyrics-Markov-Chain-Generator-
Generate Gujarati song lyrics using Markov chain model. Tokenize lyrics, create transition model, and generate new text. Utilizes NLTK library for tokenization and random selection for variability. Output represents creative continuation of original lyrics.
Library-books-managment
A C++ library management program enables users to enter personal data, borrow, extend, and return books, with menu-driven interface and error handling, promoting efficient library operations.
Movements-from-Skeleton-Data
Predict future movements from skeleton data. Utilize XGBoost classifier on time series of 3D skeleton data for tasks like fall detection or gesture recognition. Preprocess, train, evaluate, and predict for submission.
Predicting-Future-Movements-from-Skeleton-Data
This notebook tackles predicting future movements from skeleton data, crucial for applications like fall detection or gesture recognition. Leveraging deep learning, it preprocesses the dataset, trains a CNN model, evaluates its accuracy, and generates predictions for submission, all in a structured manner.
Q-Learning-for-Blackjack-in-different-environments
This repository implements Q-Learning in Blackjack, comparing it with random action selection and basic strategies. Includes experiments with various strategies, rule variations, and deck numbers to evaluate performance.
ROS-2-integration-with-MIRA-frame-work
"Integrated ROS2 into MIRA-based robot navigation, overcoming errors & hardware challenges. Improved capabilities, set stage for future enhancements. Proficient in ROS2 Nav2 stack & MIRACenter(GUI)."
Stacking-library
A powerful ensemble learning class that supports multi-layer stacking and blending models for regression tasks, with K-fold cross-validation and hold-out validation set options, for robust model performance.
Sudoku-Puzzle-Identifier-and-Solver
This repository contains a Jupyter Notebook file equipped with code for identifying the position of digits in a Sudoku puzzle and solving it. The notebook implements algorithms for Sudoku puzzle solving, offering a practical tool for enthusiasts and learners alike to explore and understand Sudoku solving techniques. 2-2 METHODS FOR BOTH TASKS.
Text-Clustering-with-Sentence-Transformers
This project demonstrates text clustering using SBERT embeddings. Includes text preprocessing, K-Means and DBSCAN clustering, and visualization of clusters. See scripts/ for detailed implementation.
Blackjack-Gameplay-Environment
A Python-based Blackjack game featuring betting, insurance, and double down options. Play against the dealer, manage your balance, and experience a realistic single-player casino environment in the terminal.
Product-Management-
A simple C++ console app to manage a list of products. Features include adding, displaying, sorting by price, and tracking total products. Limited to 10 products. Uses basic menu-based interaction.
Semantic-Data-Processing
Submission for semantic data processing module
Text-Clustering-and-Visualization
This project processes and clusters text data using BERT embeddings, K-means, and dimensionality reduction. Visualizations include t-SNE plots and word clouds. Dataset and embeddings links are provided.