Statistical Data Science

1. 👨‍🏫 Instructor & Course Logistics

• 👨‍🏫 Instructor: Mejbah Ahammad
• 🗓️ Semester: Spring Semester
• ⏰ Class Time: 8:00 PM – 10:00 PM
• 📅 Class Days: Tuesday and Friday
• 💻 Class Mode: Remote (Zoom)
• 💰 Course Fee: ৳4000
• ☎️ Contact Number: +8801874603631
• ⌚ Lessons & Time: 20 Lessons, 40 ঘন্টা 20 মিনিট total
• 📧 Email: hello@softwareintelligence.ai
• 🌐 Website: http://softwareintelligence.ai/

2. 📝 Course Description

Statistical Data Science merges:

• 🎓 Foundational Statistics (probability, distributions, hypothesis testing)
• 📂 Data Wrangling & EDA (cleaning, transformation, exploration)
• 💻 Machine Learning (regression, classification, ensemble methods, clustering)
• 🤔 Advanced Topics (dimensionality reduction, Bayesian methods, interpretability)
• 🗣️ Professional Communication (reports, dashboards, ethical & business considerations)

Students will develop an end-to-end data science pipeline, culminating in a capstone project that illustrates practical application and professional best practices.

3. 🎯 Learning Outcomes

By the end of this course, you will:

1. 🎓 Beginner-Level Skills

   • 👉 Understand fundamental probability and descriptive statistics.
   • 👉 Perform basic data loading, cleaning, and visualization in Python.

2. 📈 Intermediate-Level Skills

   • 👉 Apply hypothesis testing, regression, classification, and clustering.
   • 👉 Employ feature engineering, dimensionality reduction, and ensemble methods.

3. 💡 Advanced-Level Skills

   • 👉 Integrate Bayesian methods, neural networks, or other specialized ML techniques.
   • 👉 Assess and mitigate model bias, interpret black-box models, and use fairness frameworks.

4. 🗣️ Communication & Collaboration

   • 👉 Create professional-quality visualizations and summaries for stakeholders.
   • 👉 Collaborate effectively in teams, giving and receiving structured feedback.

4. 🏷️ Prerequisites

1. 🎓 Mathematics & Statistics

   • Basic algebra, probability, and inferential statistics (e.g., normal distribution, p-values).

2. 💻 Programming

   • Proficiency in Python (data structures, basic scripting).
   • Familiarity with NumPy, pandas, matplotlib, scikit-learn.

3. 💼 Logistics & Tools

   • Reliable internet connection for Zoom.
   • Ability to install and manage Python environments (Anaconda recommended).

5. 📚 Course Materials

A. Required Texts/Readings

1. 📗 Practical Statistics for Data Scientists by Peter Bruce & Andrew Bruce (O’Reilly).
2. 📙 An Introduction to Statistical Learning (ISL) by James, Witten, Hastie, Tibshirani (Springer).

B. Recommended & Advanced

• 📒 The Elements of Statistical Learning (ESL) by Hastie, Tibshirani, Friedman (Springer).
• 📘 Python for Data Analysis by Wes McKinney (O’Reilly).
• 📕 Bayesian Data Analysis by Gelman et al. (CRC Press).

C. Software & Tools

• 💻 Python 3.x (Anaconda Distribution)
• 📓 Jupyter Notebook (or VSCode/PyCharm)
• 🖥️ Zoom for remote sessions

6. 🗓️ 10-Week Schedule & Format

• 10 Weeks total, 20 classes (two per week).
• Each class is 2 hours: typically theory + hands-on coding/discussion.
• Participation is integral to mastering the material.

Week	Class	Level	Topic	Key Highlights
1	Class 1	Beginner	📍 Course Intro & Probability Basics	Syllabus overview, environment setup, discrete/continuous distributions
	Class 2	Beginner	📍 Data Wrangling & EDA Fundamentals	Missing values, outliers, summary stats, basic plots (pandas/seaborn)
2	Class 3	Beginner → Intermediate	📍 Statistical Inference & Hypothesis Testing	t-tests, p-values, confidence intervals, real vs. simulated data
	Class 4	Intermediate	📍 ANOVA & Experimental Design	One-way ANOVA, assumptions, multiple comparisons, A/B testing
3	Class 5	Intermediate	📍 Linear Regression (Simple & Multiple)	OLS derivation, assumptions, R-squared, residuals, coding with `sklearn`
	Class 6	Intermediate	📍 Logistic Regression & Classification Metrics	Confusion matrix, precision/recall, F1-score, ROC-AUC
4	Class 7	Intermediate	📍 Feature Engineering & Selection	Encoding (categorical, one-hot), polynomial features, feature importance
	Class 8	Intermediate	📍 Regularization (Ridge, Lasso) & Bias-Variance	Cross-validation, hyperparameter tuning, bias-variance trade-off
5	Class 9	Intermediate	📍 Dimensionality Reduction (PCA, LDA)	Eigen-decomposition, variance explained, optional t-SNE/UMAP for visualization
	Class 10	Intermediate	📍 Clustering (K-means, Hierarchical, DBSCAN)	Cluster metrics (silhouette), dendrograms, density-based approaches
6	Class 11	Intermediate	📍 Ensemble Methods (Bagging, Random Forest, Boosting)	Decision trees, random forests, AdaBoost/Gradient Boosting
	Class 12	Intermediate → Advanced	📍 Time Series or Advanced Classifier	Stationarity, ARIMA basics OR advanced algorithms (SVM, multi-class)
7	Class 13	Advanced	📍 Bayesian Methods & Probabilistic Modeling	Bayesian inference, priors/posteriors, MCMC sampling
	Class 14	Advanced	📍 Neural Networks (MLP)	Feedforward architectures, activation functions, loss functions
8	Class 15	Advanced	📍 Model Evaluation & Interpretability	Cross-validation pitfalls, LIME/SHAP, model fairness and bias mitigation
	Class 16	Advanced	📍 MLOps & Model Deployment	Flask/FastAPI, Docker, CI/CD pipelines
9	Class 17	Advanced	📍 Time Series Forecasting	ARIMA/SARIMA, trend/seasonality decomposition
	Class 18	Advanced	📍 Advanced Classification Methods	SVM tuning, XGBoost/LightGBM models
10	Class 19	Advanced	📍 Big Data & Distributed ML	Apache Spark, parallel ML processing, handling large datasets
	Class 20	Advanced	📍 Capstone Project Presentations & Future Directions	Final presentations, course wrap-up, next steps in deep learning & AI

7. 📝 Assessment & Grading

1. 📄 Weekly Assignments (40%)

   • 👉 Coding tasks, problem sets, short reflections.
   • Reinforces both conceptual and practical skills.

2. 📝 Quizzes (10%)

   • 👉 Periodic checks (announced or pop).
   • Covers fundamental stats, ML, and Python usage.

3. 💼 Capstone Project (40%)

   • 👉 Real-world data pipeline: wrangling → EDA → modeling → evaluation → presentation.
   • Teams or individuals; final presentation + written report.

4. 🤝 Participation (10%)

   • 👉 Active Zoom attendance, Q&A, breakout discussions.
   • Peer reviews and constructive feedback are essential.

🏷️ Grade Scale

• A = 90–100%
• B = 80–89%
• C = 70–79%
• D = 60–69%
• F = < 60%

8. ⚖️ Course Policies

1. 🏷️ Attendance & Engagement

   • 👉 Timely Zoom attendance, camera encouraged. Notify absences in advance.

2. 📢 Communication

   • 👉 Important announcements via email. Check daily.
   • For help or clarifications, email hello@softwareintelligence.ai.

3. ⏲️ Late Submissions

   • 👉 Potential penalties unless previously arranged.
   • Extensions granted for valid reasons (health, emergencies).

4. ⚠️ Academic Integrity

   • 👉 Plagiarism or unauthorized collaboration is prohibited.
   • Violations follow institutional policy.

5. 💻 Technical Setup

   • 👉 Ensure Python (Anaconda) is installed, Zoom stable.
   • Familiarity with version control (Git) is recommended for project work.

9. 🆘 Additional Support & Office Hours

• ⏰ Office Hours: By appointment (Zoom).
• 🎓 Extra Help: Instructor can provide supplementary resources or 1-on-1 guidance.

10. 📑 Detailed Weekly Highlights with Professional Focus

Below, each class has extra bullet points under 💼 Professional/Industry Focus to show how these concepts apply in real-world settings and build your professional toolkit.

Week 1

Class 1

• 📌 Topics: Syllabus Overview, Probability (Discrete/Continuous), Environment Setup

• 👉 Assignment:

  • Install Python libraries (NumPy, pandas, etc.).
  • Short probability exercise (theoretical + coding).

• 💼 Professional/Industry Focus:

  • Understanding basic distributions is crucial for risk assessment (finance, insurance).
  • Proper environment setup mirrors DevOps best practices in real companies.

Class 2

• 📌 Topics: Data Wrangling & EDA (Missing Values, Outliers, Basic Plots)

• 👉 Assignment:

  • Clean a small dataset; produce summary statistics and quick visualizations.

• 💼 Professional/Industry Focus:

  • Data cleaning is ~80% of real data science work: verifying data integrity is key.
  • EDA presentations often inform stakeholders about potential business decisions.

Week 2

Class 3

• 📌 Topics: Inferential Statistics (t-tests, Confidence Intervals, p-values)

• 👉 Assignment:

  • Conduct hypothesis tests on real or simulated data.
  • Present a short report on findings.

• 💼 Professional/Industry Focus:

  • Hypothesis testing underpins A/B testing in product optimization, marketing campaigns.
  • Communicating p-values/conclusions to non-technical business leaders is a vital skill.

Class 4

• 📌 Topics: ANOVA & Experimental Design (One-way ANOVA, A/B Testing)

• 👉 Assignment:

  • Compare multiple group means, interpret significance.

• 💼 Professional/Industry Focus:

  • A/B or multi-variant tests are standard in e-commerce (website design changes, user experience).
  • Solid experimental design prevents costly misinterpretations in real projects.

Week 3

Class 5

• 📌 Topics: Linear Regression (Simple & Multiple), OLS, Assumptions

• 👉 Assignment:

  • Apply multiple regression on a real dataset (e.g., housing prices).
  • Evaluate residuals, R-squared.

• 💼 Professional/Industry Focus:

  • Linear regression is the backbone for forecasting sales, pricing strategies, and resource planning.
  • Understanding assumptions is essential to avoid legal/ethical pitfalls (e.g., biased predictions in finance).

Class 6

• 📌 Topics: Logistic Regression & Classification Metrics (Precision, Recall, F1, ROC-AUC)

• 👉 Assignment:

  • Classification on Titanic-like dataset, interpret confusion matrix.

• 💼 Professional/Industry Focus:

  • Logistic regression is widely used in credit risk modeling, customer churn prediction.
  • Choosing the right metric (precision vs. recall) matters for applications like medical diagnostics vs. spam detection.

Week 4

Class 7

• 📌 Topics: Feature Engineering & Selection (Encoding, Polynomial Features, Feature Importance)

• 👉 Assignment:

  • Transform features, compare model performance with/without these transformations.

• 💼 Professional/Industry Focus:

  • Good feature engineering can drastically reduce model complexity and cost in production.
  • Feature selection helps in compliance scenarios (regulatory audits on used data fields).

Class 8

• 📌 Topics: Regularization (Ridge, Lasso) & Bias-Variance

• 👉 Assignment:

  • Tune alpha in Ridge/Lasso; compare error rates.

• 💼 Professional/Industry Focus:

  • Regularization is crucial for financial forecasting or marketing analytics where overfitting can be expensive.
  • Cross-validation is an industry standard for robust model validation before deployment.

Week 5

Class 9

• 📌 Topics: Dimensionality Reduction (PCA, LDA, Optional t-SNE)

• 👉 Assignment:

  • PCA on a high-dimensional dataset; interpret principal components.

• 💼 Professional/Industry Focus:

  • PCA is essential in high-dimensional scenarios (e.g., genetics data, sensor data).
  • Reducing features can improve processing speed and help in real-time applications.

Class 10

• 📌 Topics: Clustering (K-means, Hierarchical, DBSCAN)

• 👉 Assignment:

  • Apply at least two clustering methods; evaluate with silhouette score.

• 💼 Professional/Industry Focus:

  • Clustering is pivotal for customer segmentation and market research.
  • Hierarchical clustering often used in gene expression analysis or text analytics.

Week 6

Class 11

• 📌 Topics: Ensemble Methods (Bagging, Random Forest, Boosting)

• 👉 Assignment:

  • Compare random forest & gradient boosting on a classification or regression dataset.

• 💼 Professional/Industry Focus:

  • Ensemble methods dominate Kaggle competitions and are widely used in finance (fraud detection) and healthcare (diagnostics).
  • Random forests offer interpretability advantages in regulatory contexts compared to black-box models.

Class 12

• 📌 Topics: Time Series or Advanced Classifier (Choose Focus)

  • Option A: Time Series – Stationarity, ARIMA, seasonal patterns
  • Option B: Advanced Classification – SVM, multi-class strategies

• 👉 Assignment:

  • Forecast a simple time series OR tune an SVM for a multi-class dataset.

• 💼 Professional/Industry Focus:

  • Time series forecasting is critical in inventory management, financial trading.
  • Advanced classifiers (SVM) are used for image classification, bioinformatics.

Week 7

Class 13

• 📌 Topics: Bayesian Methods & Probabilistic Modeling (Priors, Posterior, MCMC Intro)

• 👉 Assignment:

  • Implement Bayesian updates on a small dataset; compare to frequentist approach.

• 💼 Professional/Industry Focus:

  • Bayesian inference is key in medical trials, market research (incorporating prior knowledge).
  • MCMC methods are used in complex risk modeling (e.g., insurance, actuarial science).

Class 14

• 📌 Topics: Neural Networks (MLP) – Activation Functions, Feedforward Architecture

• 👉 Assignment:

  • Train a small MLP on a classification dataset (e.g., MNIST or tabular).

• 💼 Professional/Industry Focus:

  • Neural nets power computer vision (e-commerce product tagging) and NLP (chatbots, sentiment).
  • Balancing data requirements vs. model complexity is crucial for cost and performance in production.

Week 8

Class 15

• 📌 Topics: Model Evaluation & Interpretability (CV pitfalls, LIME/SHAP, Fairness)

• 👉 Assignment:

  • Apply an interpretability tool to a trained model; analyze bias or feature impact.

• 💼 Professional/Industry Focus:

  • Many industries (finance, healthcare) require interpretability to comply with regulations.
  • Tools like SHAP help build trust with clients and executives.

Class 16

• 📌 Topics: MLOps & Model Deployment (Flask/FastAPI, Docker, CI/CD)

• 👉 Assignment:

  • Containerize a model and deploy a simple API locally or on a cloud platform.

• 💼 Professional/Industry Focus:

  • Productionizing models is a core skill for data scientists in tech companies.
  • Docker/CI-CD ensures reproducibility and quick iteration in enterprise solutions.

Week 9

Class 17

• 📌 Topics: Capstone Project Workshop (Data Debugging, Methodology Refinement)

• 👉 Assignment:

  • Submit capstone progress outline or preliminary code.

• 💼 Professional/Industry Focus:

  • Project management (timeline, scope) aligns with agile methodologies used in industry.
  • Peer feedback mimics code reviews or project stand-ups in real teams.

Class 18

• 📌 Topics: Capstone Presentations (Part 1)

• 👉 Deliverable:

  • Live demos, peer Q&A, instructor critique.

• 💼 Professional/Industry Focus:

  • Presentation skills are essential when pitching data insights to C-level executives or non-tech stakeholders.
  • Showcasing end-to-end solutions fosters a consultative approach to data problems.

Week 10

Class 19

• 📌 Topics: Capstone Presentations (Part 2)

• 👉 Deliverable:

  • Remaining presentations, advanced discussion of methodology.

• 💼 Professional/Industry Focus:

  • Final demos reflect client-facing scenarios in consulting or internal data science teams.
  • Handling tough Q&A showcases confidence and readiness for industry interviews or stakeholder sessions.

Class 20

• 📌 Topics: Course Wrap-Up & Future Directions (Big Data, Deep Learning, Specialized Domains)

• 👉 Assignment:

  • Submit final capstone code/report.
  • Complete course evaluation survey.

• 💼 Professional/Industry Focus:

  • Understanding next steps (Spark/big data, advanced deep learning) is essential for scaling solutions.
  • Networking, continuous learning, and professional development keep data scientists at the cutting edge.

✅ Final Note

Welcome to Statistical Data Science! Over the next 10 weeks, we will bridge fundamental statistics and modern data science practices, with each class enriched by professional insights. Keep these key points in mind:

• Practice regularly and experiment with different datasets.
• Communicate your work effectively—technical mastery + clarity = real-world impact.
• Collaborate and ask questions—learning from peers is invaluable.

We look forward to a dynamic and career-focused semester together!

👨‍🏫 Instructor: Mejbah Ahammad
📧 Email: hello@softwareintelligence.ai
☎️ Phone: +8801874603631
🌐 Website: http://softwareintelligence.ai/

(C) 2025 Software Intelligence & Intelligence Academy – All Rights Reserved.