- Experienced data scientist delivering multi-million-dollar projects for global companies.
- Adept at handling complex, cross-functional datasets to identify and extract insights.
- Proven track record of data-driven decision-making and predictive modeling.
- Reliable team player with excellent communication and analytical skills.
- Passionate about innovation and driving sustainability goals.
With a strong emphasis on machine learning, my portfolio encompasses a range of end-to-end projects in NLP, regression, classification, and deep learning. Additionally, my projects are underpinned by comprehensive exploratory data analysis techniques that help me to identify key trends and patterns in data sets. By combining these skills, I am able to deliver high-quality data-driven solutions that drive business outcomes.
Methodology: Binary classification
Algorithms: LogisticRegression, MNB, RF, SVC
Performance: Accuracy=85%, ROC AUC=0.92
Summary: Developed a classifier model to create an ad campaign strategy for appropriate subreddits.
Methodology: Unstructured Data-Cluster Analysis
Algorithms: LDA, GSDMM
Performance: Identified 50 clusters
Summary: Identified several clusters to summarize the context of tweeting activity on Twitter
Methodology: Image Classification
Algorithms: CNN
Performance: log-loss=0.73, accuracy=0.7
Summary: Developed an image classification model to detect driver behavior
Methodology: EDA using Python, Pandas & MatplotLib
Algorithms: Data Analysis
Performance: Insights on game play
Summary: Developed several visualizations to depict the journey of Kobe Bryant's professional career.
Methodology: Regression
Algorithms: Linear Regression, Regularization Using Ridge & Lasso
Performance: RMSE score = 18,000
Summary: Predictive Modeling for Sales Price: Unveiling Influential Predictors through Regression Analysis