Milan Dora

Stock Price Prediction Project

• Built a machine learning model to predict SBI stock prices using historical daily data (Jan 2020–Apr 2025), aiming to capture complex market trends for accurate forecasting.
• Employed feature engineering with technical indicators (WMA, RSI, MACD, etc.), multicollinearity analysis (VIF), Yeo-Johnson transformation, StandardScaler, and TimeSeriesSplit for robust preprocessing and evaluation.
• Developed and tuned Support Vector Regression (SVR), Random Forest, Multi-Layer Perceptron (MLP), and Artificial Neural Network (ANN) using GridSearchCV.

Power Demand Forecasting Project

• Developed a machine learning model to predict hourly power demand in Delhi using 2023 data, addressing challenges from seasonal fluctuations, weather, and the Duck Curve for grid stability and resource planning.
• Applied feature engineering (lagged features, season encoding, cyclic hour features), Yeo-Johnson transformation, PCA for dimensionality reduction (95% variance, 9 components), RobustScaler for target scaling, and TimeSeriesSplit for evaluation.
• Trained and tuned Support Vector Regression (SVR), Random Forest, and Long Short-Term Memory (LSTM) models using GridSearchCV for hyperparameter optimization.

Movie Recommendation System with Emotion-Based Filtering (https://movierecomendationemotions-ieqkjhbo9zpqfzpzfcmlhc.streamlit.app/)

• Developed a web-based movie recommendation system to suggest films tailored to users’ emotional states or preferences, enhancing personalized content discovery using a Streamlit interface.
• Implemented content-based or collaborative filtering, likely with feature engineering (e.g., movie genres, user ratings, or sentiment analysis), data preprocessing (using Pandas/NumPy), and Streamlit for an interactive UI.
• Utilized recommendation algorithms (e.g., cosine similarity, matrix factorization, or clustering) to match movies with user inputs, possibly integrating NLP for emotion-based filtering.

Breast Cancer Classification Web App (https://breastcancer-pbutao88htpvhsnk95txz4.streamlit.app/)

• Developed a Streamlit web application to classify breast tumors as malignant or benign, leveraging machine learning to support early diagnosis and medical decision-making.
• Utilized data preprocessing (e.g., StandardScaler, label encoding), exploratory data analysis (e.g., scatter plots, heatmaps).
• Implemented classification algorithms such as Logistic Regression, Random Forest, or K-Nearest Neighbors (KNN), optimized via hyperparameter tuning (e.g., GridSearchCV).

• Worked on multiple projects like Smart Supplies Forecasting, Smart Supplies Planning, Planning Manufacturing and Development applications.
• Collaborated with cross-functional teams to establish quality standards, ensuring alignment with organizational goals and objectives.
• Improved product quality by implementing rigorous testing protocols and identifying areas for improvement.
• Communicated between product owners and developers to clarify features.
• Optimized workflow efficiency by automating repetitive testing tasks, freeing up resources for higher-value activities.
• Operated under Agile and Scrum frameworks to complete releases and well-organized sprints.
• Evaluated function, performance and design compliance of every product against design standards and customer needs.

• Designing test cases from requirements by covering all test conditions and eliminating redundancy and duplications.
• Developed and maintained quality assurance procedure documentation.
• Manual Testing of APIs using Postman tool, SQL database testing, backend testing and handling team with Manager.