Manish Kumar Mahto
Hi , I'm Manish
I am a
Entry-Level Data Scientist | Python • ML • Deep Learning • LangChain & RAG Systems | Showcasing my projects and learning journey.
Entry-Level Data Scientist | Python • ML • Deep Learning • LangChain & RAG Systems | Showcasing my projects and learning journey.
Introduction
I'm a passionate Data Science and Machine Learning enthusiast. With a focus on data analysis, deep learning, and NLP, I aim to solve real-world problems through data-driven solutions.
Some of my key projects include a Movie Recommendation System 🎥, Quora Question Classification 📝, and Churn Prediction using ANN 💡. I apply technologies like XGBoost, ANN, CNN, and LSTM to build innovative models.
Always eager to expand my skills, I am excited to explore opportunities in AI, Deep Learning, and Data Science. Let's connect and discuss how we can leverage data for impactful results! 👨💻
Technical Skills
Recent Work
A Fast API web app where patient management system implemented along with claim Prediction Machine Learning Algorithm.
A deep learning-based Seq2Seq translation model for converting English to Hindi and Hindi to English using Encoder Decoder Architecture Research Paper.
Forecasted TCS stock's next-day closing price using a hybrid LSTM + Dense model and optimized Random Forest for multi-day predictions with up to 97.5% R² accuracy.
Achieved 92.25% test accuracy and 93.75% validation accuracy on GPU using TensorFlow with a fine-tuned VGG19 model.
Classified Quora duplicate question pairs using feature engineering and Random Forest, achieving 78% accuracy.
Built a Content-Based Movie Rec. System with 75% memory efficiency and an interactive Streamlit app using Python and Scikit-learn.
Streamlit dashboard with K-Means (5 clusters) for retail segmentation, interactive filters, and visualizations. Uses StandardScaler and Seaborn/Matplotlib.
Recommends books based on collaborative filtering, with top 5 similar book suggestions and a display of the top 50 books by average rating.
Developed an ANN model with TensorFlow to predict bank customer churn, achieving high accuracy of 86%. Deployed it in a Streamlit app for real-time use.
Developed a Mobile Price Predictor achieving 81.57% R² with XGBoost, leveraging extensive data preprocessing and feature engineering.
I created a machine learning model to predict car prices using Python and Scikit-learn.
Conducted a Diwali Sales Analysis using Python, uncovering key trends and insights through data cleaning, EDA, and visualizations with NumPy, Pandas, Seaborn, and Matplotlib.
Currently undertaking an 8-week SQL challenge to master advanced querying techniques and data manipulation through real-world scenarios.
