Rooban J

Experience

• 

  Wheels India Limited

  Oct 2021 - Dec 2022

  Graduate Engineering Trainee
• 

  AlmaBetter

  Feb 2023 - now

  Data Science Trainee

  Involved in learning of all essential data analysis requirement starting from Python libraries (Numpy, Pandas, Matplotlib, Seaborn), SQL functions & queries, Advanced Excel Concepts, Data Visualization via Tableau and Topics on Machine learning algorithms (Regression, Classification, NLP, etc.,)List of Projects which I have involved during my tenure.,--EDA PROJECT Extracted insights from hotel booking dataset Conducted Exploratory Data Analysis (EDA) on a hotel booking dataset. During the analysis, we employed various data transformation techniques to process the data and extract insights. Additionally, we utilized multiple visualization methods to identify relationships between features, aiming to uncover the most influential factors impacting profits.--REGRESSION MODEL (Supervised Learning)Extracted insights from bike sharing demand dataset Developed Supervised Machine Learning Models to predict the count of bikes required to meet demand for bike sharing, and determined the significance of features influencing the project outcome. The models employed in this project include Linear Regression, Lasso Regression, Ridge Regression, Elastic Net Regression, Decision Tree, Random Forest, and Gradient Booster. Additionally, we applied cross-validation techniques and performed hyperparameter tuning to optimize model performance. Evaluation metrics were used to validate the models.--CLASSIFICATION PROJECT (Supervised Learning)Extracted insights dataset from mobile industry Developed Supervised Machine Learning Models to identify influential factors impacting the price of mobile phones. In this project, we experimented with Logistic Regression, XGBoost, and Random Forest Classifier algorithms. Our findings suggest that logistic regression and XGBoost, with hyperparameter tuning, delivered the most accurate predictions for the price range of mobile phones. Model explainability tools were utilized to identify the most important features for prediction. Show less

• 

  Aparajitha Corporate Services Private Limited

  Jul 2023 - Apr 2024

  Executive Personnel
• 

  CodSoft

  Apr 2024 - now

  Data Science Intern

  During my internship, I had the opportunity to work with real-world datasets and build models to achieve specific outcomes. The insights and skills gained during this experience have greatly enhanced my understanding of data science and its practical applications. Some of the projects I worked on during my internship include:Project 1 - TITANIC SURVIVAL PREDICTION This project involves real-time data concerning the Titanic voyage and the accident, including survival rates. The dataset provided demanded a strong sense of visualization thinking and techniques. I implemented data pre-processing techniques to develop a predictive model and evaluated the performance through evaluation metrics. The learnings and insights gained from this project have significantly deepened my understanding of data science and its real-world applications.Project 2 - MOVIE RATING PREDICTION In this project, I developed a machine learning model for predicting movie ratings using the IMDB dataset from Kaggle. I implemented the Linear Regression technique for prediction. Through preprocessing, visualization, and analysis, the resulting solution accurately predicts ratings. Finally, we evaluated our model's performance using evaluation metrics. This project has not only enhanced my learning but also provided valuable experience in handling data ambiguity.Project 3 - IRIS FLOWER CLASSIFICATION The Iris dataset is renowned for its comprehensive set of parameters, which provides a captivating opportunity for analysis. The aim of this project is to construct a machine learning model for predicting the category of Iris flowers. Utilizing Python and Scikit-Learn, I implemented multiple classification algorithms. After establishing the accuracy of the models, it was observed that the Random Forest Classifier algorithm demonstrated higher performance efficiency on the training data. However, the Logistic Regression model exhibited superior accuracy on both the training and test datasets. Show less