Big Data and Predictive Analytics

This foundational course introduces students to the core principles of data science while building practical programming skills in Python and R—two of the most widely used languages in analytics. Students explore variables, data types, control structures, functions, libraries (such as NumPy, pandas, and ggplot2 which is a powerful data visualization package in R), and scripting techniques that are essential for handling and analyzing data. Beyond syntax, students gain a strong understanding of the problem-solving approaches, data flow logic, and toolsets commonly used in the data science workflow. Expanded topics include Jupyter Notebooks, RStudio, integrated development environments, and version-controlled scripting. Emphasis is also placed on reproducibility, debugging, and interpreting code used in predictive analytics applications.

This course provides essential techniques for cleaning, transforming, and preparing raw data for analysis. Students learn to identify and resolve data quality issues such as missing values, outliers, inconsistent formats, and duplicates using Python’s pandas, R’s dplyr, and SQL queries. Emphasis is placed on creating reproducible workflows, data pipelines, and using regular expressions for text cleaning. Expanded content includes data type conversions, handling date/time formats, flattening nested structures (JSON/XML), and working with unstructured and semistructured datasets. Students complete labs simulating real-world messy datasets from health, finance, and social media domains.

Students develop a strong grounding in statistical reasoning and probabilistic thinking, key to effective data interpretation and modeling. Topics include descriptive statistics, distributions (normal, binomial, Poisson), central tendency, dispersion, correlation, and regression. Inferential statistics such as hypothesis testing, p-values, confidence intervals, and sampling techniques are also covered. Expanded content introduces multivariate analysis, ANOVA (Analysis of Variants), Bayesian concepts, and statistical assumptions in predictive modeling. The course uses hands-on labs and real-world datasets to link statistical tools to business problems and decision-making processes.

Students explore the core supervised learning methods used in predictive analytics. They build and evaluate models using linear regression, logistic regression, decision trees, random forests, and support vector machines. Students learn about training/testing data splits, cross-validation, performance metrics (precision, recall, F1 score, AUC), and model selection strategies. Expanded topics include feature engineering, hyperparameter tuning, model bias/variance trade-offs, and handling imbalanced datasets. Tools such as scikit-learn and caret are used extensively, and labs emphasize iterative development and deployment of accurate predictive models.

This course introduces students to algorithms that identify hidden patterns or groupings in unlabeled data. Students learn to implement K-means, DBSCAN, and hierarchical clustering, and evaluate clustering outcomes using silhouette scores and distance metrics. Topics also include dimensionality reduction (t-distributed stochastic neighbour embedding), market segmentation, anomaly detection, and recommender systems. Expanded content emphasizes real-world applications in fraud detection, customer profiling, and exploratory data analysis. Students gain experience analyzing large datasets and visualizing clusters for actionable insights.

Students gain hands-on experience with distributed computing frameworks essential for processing massive datasets. They learn the architecture and components of Hadoop (HDFS, MapReduce, YARN) and Apache Spark for inmemory processing and real-time analytics. Expanded topics include data ingestion tools (Sqoop, Flume), Hive for SQL-like queries on Hadoop, Spark SQL, and Spark MLlib for machine learning at scale. Students also deploy Spark jobs in local and cluster environments and explore best practices in scaling analytical workloads efficiently.

This course introduces students to cloud environments (AWS, Azure, or Google Cloud) and their application in big data analytics. Students learn to use cloud-based tools for data storage (S3, Blob Storage), compute services (EC2, Databricks, Lambda), and managed databases. Expanded topics include configuring cloud environments for analytics pipelines, data warehousing (Redshift, BigQuery), and serverless architecture. Students gain exposure to real-world workflows, including ingestion, processing, visualization, and deploying machine learning models in the cloud.

Students learn to transform analytical results into clear, engaging dashboards and reports using BI tools such as Tableau, Power BI, or Looker. Topics include data blending, filtering, aggregation, calculated fields, interactive visualizations, and storytelling techniques. Expanded coverage includes KPI development, trend analysis, geospatial visualization, and real-time dashboarding. Emphasis is placed on aligning visual output with business goals and tailoring presentations to both technical and non-technical audiences.

Students build data pipelines that automate the extraction, transformation, and loading (ETL) of data across systems. Tools such as Apache Airflow, Talend, and SQL are used to build workflows for batch and stream processing. Topics include scheduling, job orchestration, logging, error handling, and pipeline optimization. Expanded content includes designing schema-aware pipelines, managing dependencies, implementing data quality checks, and integrating pipelines with cloud services and big data frameworks. The course simulates enterprise-scale data engineering workflows and architecture.

This course focuses on analyzing data indexed in time order to identify trends, seasonality, and cycles. Students apply techniques such as Auto Regressive Integrated Moving Average (ARIMA), exponential smoothing, and seasonal decomposition using statistical packages in R or Python. Expanded content includes stationarity testing, autocorrelation, model diagnostics, rolling forecasts, and applying time series to business scenarios such as sales forecasting, patient admission modeling, and energy usage prediction.

Students explore the frameworks and regulations that ensure responsible data handling, including GDPR, HIPAA, and PIPEDA. Topics include anonymization, consent, privacy-preserving analytics, audit trails, data classification, and access control. Expanded topics include algorithmic fairness, ethical AI, explainability, and bias mitigation in modeling. Case studies reinforce ethical dilemmas and best practices in balancing innovation with public trust and legal compliance.

Students synthesize their learning in a comprehensive project that involves problem definition, data acquisition, cleaning, modeling, and presentation. Projects may simulate scenarios in sectors such as healthcare (e.g., patient outcome prediction), finance (e.g., risk modeling), or marketing (e.g., customer churn analysis). Expanded expectations include deploying a model via a cloud platform, creating a data visualization dashboard, and presenting a full project report with code, methodology, results, and recommendations. Students develop a portfolio-ready artifact that demonstrates their readiness for analytics roles.