Welcome to Mustaaq's Portfolio

Data Science | Data Analytics | Machine Learning

Passionate about data, algorithms, and building scalable solutions.

Explore Projects

🙋‍♂️About Me

🚀 Data Scientist | Machine Learning & Cloud Enthusiast

Hi, I’m Mustaaq Ahmed Khan, a driven and detail-oriented Graduate Student pursuing a Master of Science in Computer Science at the University of Bridgeport. With a strong passion for transforming raw data into actionable insights, I specialize in leveraging cutting-edge tools and technologies to solve complex real-world challenges. My expertise spans programming languages like Python, SQL, R, and C++, as well as frameworks such as Scikit-learn, TensorFlow, Keras, and XGBoost, enabling me to build robust machine learning models and deliver impactful solutions.

As a Data Science Intern at Delta Sigma Technologies , I optimized data processing workflows, reducing processing time by 90% through efficient algorithms and parallel computing. I analyzed over 500K data points weekly to enhance predictive analytics models, improving decision-making processes by 30% and delivering findings with 90% accuracy. This experience honed my ability to translate data-driven insights into tangible business outcomes.

I recently completed a Virtual Data Analyst Internship Certification from Trainity , where I worked on diverse projects such as Instagram User Analytics, Operational & Metric Analytics, Hiring Process Analytics, IMDB Movie Analysis, Bank Loan Case Studies, Impact of Car Features, and ABC Call Volume Trends. These projects strengthened my proficiency in Excel, SQL, Python, Power BI, and Tableau while sharpening my skills in exploratory data analysis (EDA), data visualization, and deriving actionable business insights. Through this certification, I developed a keen eye for identifying patterns, optimizing workflows, and creating scalable solutions that drive efficiency and accuracy in data-driven decision-making.

I am currently exploring Azure ML and MLOps , preparing for the DP-100 certification , with the goal of bridging the gap between machine learning development and deployment. My ultimate aspiration is to leverage data science and AI to drive strategic decision-making in high-stakes environments, such as Formula 1 strategy engineering, where precision and speed are paramount.

With a solid foundation in data analytics , cloud computing , and machine learning , coupled with hands-on experience in tools like Power BI, Tableau, and Excel, I am excited to push the boundaries of what’s possible with data. I thrive on solving real-world problems through innovative approaches and am eager to contribute to transformative projects that harness the power of AI and data science. 🌟

Let’s connect and explore how we can turn data into decisions! 🚀

🛠Technical Skills

💻 Programming & Scripting Languages:

Python (Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn, TensorFlow, PyTorch)
SQL (MySQL, PostgreSQL, SQL Server, BigQuery)
R (ggplot2, dplyr, tidyr, caret)
C, C++ (Data Structures, Algorithms)
Bash & Shell Scripting (Linux, command-line automation)
HTML, CSS, JavaScript (for basic web-based data visualization)

📊 Data Analytics & Business Intelligence:

Excel (Pivot Tables, VLOOKUP, Macros, Power Query, Data Cleaning)
Power BI (DAX, Power Query, Report Building, Data Modeling)
Tableau (Dashboards, Data Blending, Calculated Fields)
Google Data Studio (Visualization & Dashboarding)
ETL Pipelines (Extract, Transform, Load using SQL, Python)
A/B Testing & Experimentation

📈 Machine Learning & AI:

Supervised Learning (Linear Regression, Logistic Regression, Decision Trees, Random Forests, XGBoost, SVM)
Unsupervised Learning (Clustering - K-Means, Hierarchical, DBSCAN; PCA, t-SNE)
Deep Learning (ANN, CNN, RNN, LSTMs, Transformers - TensorFlow, PyTorch)
NLP (Text Preprocessing, Sentiment Analysis, Named Entity Recognition, BERT)

🧑‍💻 Software & Tools:

Git & GitHub (Version Control)
Jupyter Notebook, Google Colab
VS Code, PyCharm, RStudio
Matlab (for mathematical computing)
Hugging Face (NLP Pre-trained Models)

💡 Projects

🔍Social Media Sentiment Analysis

Conducted Social Media Sentiment Analysis using Natural Language Processing (NLP) to evaluate brand reputation, achieving 89% accuracy in sentiment classification.

Analyzed social media data to identify positive, negative, and neutral sentiments, providing insights into customer perceptions and brand performance.

Visualized sentiment trends and key themes to highlight areas of strength and opportunities for improvement in brand communication and customer engagement.

Enabled data-driven decision-making for marketing and PR teams by delivering actionable recommendations to enhance brand reputation and customer satisfaction.

🔗Github Repository: Social Media Sentiment Analysis

🚕NYC Taxi Fare Prediction

Built a machine learning model to predict NYC taxi fares using features like pickup/dropoff coordinates, passenger counts, and time-based variables, achieving strong predictive accuracy with Random Forest

Performed data cleaning, feature engineering (e.g., trip distance, hour of day), and exploratory data analysis to uncover patterns in fare distributions and geographical trends.

Evaluated models using metrics like Mean Squared Error (MSE) and R-squared, with Random Forest outperforming simpler models like Linear Regression.

Identified opportunities to improve predictions by incorporating external data (e.g., weather, traffic) and experimenting with advanced models like XGBoost or deep learning.

🔗Github Repository: NYC Taxi Fare Prediction

📊 Risk Analytics for Loan Default

Conducted Exploratory Data Analysis (EDA) on a loan application dataset to identify patterns influencing loan defaults, focusing on customer and loan attributes to improve decision-making for loan approvals.

Handled missing data and outliers using Excel functions like ISBLANK, QUARTILE, and conditional formatting, ensuring data quality while visualizing distributions and anomalies through bar charts, box plots, and scatter plots.

Analyzed data imbalance in loan repayment scenarios (e.g., payment difficulties vs. on-time payments) and performed univariate, segmented, and bivariate analyses to uncover key drivers of default risk using pivot tables, histograms, and correlation matrices.

Identified top correlations between variables (e.g., income, loan amount) and repayment behavior for different scenarios, providing actionable insights to mitigate financial risks and optimize lending strategies.

🔗Drive Link: Bank Loan Case Study

🎬IMDB Movie Analysis

Conducted an IMDB Movie Analysis to uncover factors influencing movie success, defined by high IMDB ratings, providing actionable insights for producers, directors, and investors.

Analyzed the impact of genres, duration, language, directors, and budgets on movie ratings using descriptive statistics (mean, median, variance) and visualizations like scatter plots, bar charts, and trendlines in Excel.

Explored relationships such as the correlation between movie budgets and gross earnings , identifying high-profit-margin films, and assessed how directors and languages influence ratings through percentile analysis and comparative statistics.

Applied the Five Whys technique to dig deeper into trends, such as why higher-budget movies tend to succeed, revealing root causes like enhanced production quality and viewer satisfaction driving positive reviews.

🔗Drive Link: IMDB Movie Analysis

🏎️🔍Formula 1 2024 Season Analysis: Performance Trends & Predictive Insights

Drivers & Teams Performance : Points progression, qualifying vs. race performances, lap time trends, position changes, consistency metrics, and teammate comparisons.

Track & Strategy Insights : Track-based strengths, finishing position trends, tire strategy impacts, and pole lap comparisons (sector-wise, throttle usage, speed differentials).

Race Dynamics :Safety car influence, pit stop efficiency, weather conditions, and their effects on race results.

Using data from FastF1, the official F1 website, and other sources, I engineered a dataset incorporating features like tire compounds, weather conditions, safety car periods, and pit stop performance.

For predictive modeling, I built an XGBoost regression model with extensive hyperparameter tuning, achieving an R² score of 0.96. The model incorporates engineered features such as relative lap time, qualifying impact, and pit stop efficiency to predict final race positions.

This analysis not only highlights the competitive landscape of the 2024 season but also provides data-driven insights into key race determinants, helping to understand what separates podium finishers from the rest of the grid.

🔗Github Repository: Formula 1 2024 Season Analysis

📈⚡Operation Analytics & Metric Spike Investigation

Conducted Operational Analytics to analyze end-to-end operations and investigate metric spikes, such as drops in user engagement or sales, using advanced SQL queries on datasets like job_data, users, and events.

Calculated key operational metrics , including jobs reviewed per hour, 7-day rolling averages of throughput, and language share percentages, while identifying duplicate rows in the data to ensure accuracy.

Analyzed user engagement, growth, and retention through weekly cohort analysis and device-specific engagement metrics, providing actionable insights to improve operational efficiency and user experience.

Investigated email engagement and other event-based metrics to uncover trends and anomalies, helping teams understand sudden changes and optimize strategies for better performance.

🔗Drive Link: Operation Analytics

🧑‍💼Hiring Process Analytics

Conducted Hiring Process Analytics for a multinational company to evaluate trends in recruitment, including gender distribution, salary patterns, and departmental roles, providing insights to optimize hiring decisions..

Analyzed gender distribution of hires and calculated the average salary offered using Excel functions, while creating salary class intervals to visualize salary ranges and identify distribution trends.

Used visualizations like pie charts and bar graphs to represent the proportion of employees across departments and position tiers, offering clarity on workforce composition and hierarchy.

Handled missing data , combined relevant categories for streamlined analysis, and addressed outliers to ensure accurate and meaningful conclusions about the hiring process.

🔗Drive Link: Hiring Analytics