Program Outline

Data Science & Machine Learning

Career Path
Batch 01

Course Instructors

Md. Aksadur Rahman

Adjunct Lecturer,
United International University (UIU)
Data Science Consultant,
Bdjobs.com Ltd
Lead Trainer of EDGE Projects,
ICT Division, BCC
 Sabbir Hossain Rossi
Senior Business Intelligence Analyst,
German-based Retail Company
&
Former Data Scientist Team Lead,
Celloscope & DOER Services Limited

Araf Mustavi
Analyst, IDT Operations
BAT Bangladesh
Ex - Machine Learning Engineer,
ACI Limited
 Tools & Technology

Database Fundamentals & SQL (BigQuery) for Data

Scientists,
Python Fundamentals,
Statistics with Python for Data Scientists,
Machine Learning with Python,
MLOps,
Deep Learning with Python

***
How to Effectively Apply for a Data Scientist//ML Engineer Position?
Interview Skill Development
Freelancing Guidance
 Python Fundamentals

Week 01
● Class 1: Basics of Python Programming Language
❖ Introduction to Python Programming Language.
❖ Introduction to Google Colab & Jupyter Notebook for Writing and Executing
Python Codes.
❖ Getting used to the User Interface (UI) of Jupyter Notebook/Google Colab.
❖ Python Data Types: Numeric, Strings, Boolean, List, Dictionary, Tuple, Set,
None.
❖ Iterables & Mutability; List vs. Tuples.

● Class 2: Variables, Type Conversion, Operators & Methods

❖ Assigning & Overwriting Variables; Naming Convention for Variables.
❖ Type Conversion.
❖ Operators: Arithmetic & Logical/Comparison.
❖ Numeric & String (Text) Functions.
❖ String Indexing & Slicing.
❖ String Methods.
❖ F String.

● Class 3: Conditional Statement, User Defined Functions & Iteration

❖ IF & ELIF Statements.
❖ How to Define/Create a Function in Python?
❖ How to Create a Function with Parameter/Argument(s)? Function within a
Function.
❖ Use Conditional Statements and Functions Together.
❖ Using FOR & WHILE Loops.
❖ Combining Conditional Statements & Loops.
❖ Combining Conditional Statements, Functions & Loops.
❖ How to Iterate over a Dictionaries.

*** By the end of the Python Fundamentals module, students will have built a
solid foundation in core Python programming concepts and practices. This is
critical for anyone pursuing a career in data science or as a machine learning (ML)
engineer, as Python is the primary language used in these fields.
Outcomes:

❖ Mastery of Data Types.

❖ Understanding Python Basics.
❖ Variables and Type Operations.
❖ String Handling and Methods.
❖ Control Flow and Logic.
❖ Functions and Modular Programming.
❖ Iterative Processing.

Database Fundamentals & SQL for Data Scientists

Week 02
● Class 4: Fundamentals of DBMS-I
❖ What is a Relational Database Management System?
❖ ACID Property.
❖ How is data stored in a relational database?
❖ Concept of Normalization.
❖ OLTP vs. OLAP.
❖ Database vs. Data Warehouse vs. Data Lake/Lakehouse.
❖ What is a NoSQL Database and BASE Property?
❖ Difference between Relational and NoSQL Database.
❖ Which one should you choose in which case?
❖ For BI Solution, what should you choose to store your data?
❖ What is Snowflake and what advantage does it provide? (Micro-Partitioning)

● Class 5: Fundamentals of DBMS-II

❖ Installing MySQL Locally.
❖ MySQL Workbench UI Tour.
❖ Concepts of Database, Schema, Table and Fields.
❖ Types of SQL Commands: DDL, DML, DCL, TCL, DQL.
❖ What are the Commands under each of these categories and what does each of
these commands do?
❖ As a Data Analyst & Data Scientist which Commands do you only need to
Master?
❖ Constraints: Primary Key, Foreign Key, Not Null, Unique, Check, Default.
❖ Data Types: DATE, DATETIME, VARCHAR, INT, NUMBER, FLOAT, Auto
Increment.
 ● Class 6: Designing a Relational Database in MySQL & Bulk Insertion
❖ Designing and Creating Database, Schema, Tables.
❖ Masterclass on DDL, DML, TCL & DQL Commands/Statements.
❖ Bulk Insert of Data in MySQL.
❖ Error Handling during Bulk Insertion.

Week 03
● Class 7: Exploratory Data Analysis using SQL
❖ Basic Queries: SELECT, FROM, WHERE, LIKE, ILIKE, IN, DISTINCT,
BETWEEN, GROUP BY, ORDER BY, LIMIT, OFFSET, ALIAS.
❖ Aggregate Functions: COUNT, SUM, AVG, MIN, MAX.
❖ Difference between WHERE and HAVING Clause.
❖ Some built-in Functions: EXTRACT, DATE_PART, TO_DATE, TO_CHAR.
❖ CASTING, SUBSTRING, POSITION, COALESCE, NULLIF.
❖ Removing Duplicates.

● Class 8: JOINING, UNION & CASE WHEN

❖ SQL Join: Left, Right, Inner & Full Join.
❖ Be aware of Cross Join!
❖ UNION.
❖ SQL Code Order of Execution.
❖ Wide use of CASE WHEN Statement during data Cleaning, Analysis & Feature
Engineering.

● Class 9: Subqueries, CTEs & RFM Segmentation

❖ Use of Subqueries & CTEs in long SQL Codes.
❖ Segmenting Customers based on their Recency, Frequency & Monetary value
using SQL.
❖ Between Subqueries & CTEs one is more efficient?

Week 04
● Class 10: Window Functions
❖ Window Functions, the most widely used SQL Commands used by Data Analysts
& Data Scientists.
 ❖ Window Functions: RANK, DENSE_RANK, ROW_NUMBER, LEAD, LAG,
FIRST VALUE, AGGREGATE WINDOW FUNCTION, FRAME
SPECIFICATION, WINDOW CHAINING.

● Class 11: Cohort Analysis & Creating VIEWS

❖ Cohort Analysis using Transactional Customer Data.
❖ Customer Lifetime Value, Retention, and Churn Rate Calculation.
❖ Use Case of Views; How to Create Views?

*** By the end of this module, Students would be able to confidently perform any
sort of Data Analysis and Reporting for different Departments in any Organisations
using SQL. They would be well-equipped to understand any complex queries;
validating data and help any Business by generating important KPIs.

They would also be able to help Businesses by Performing Complex Analyses like
Cohort and RFM Segmentation. They would understand the importance of
VIEWS, Stored Procedures, Triggers which would help them when they would be
working on Python/Power BI/Tableau, etc. with Database/Data Warehouse as Data
Sources.

Statistics with Python for Data Science

Week 04
● Class 12: Descriptive Statistics
❖ Types of Variables.
❖ Level of Measurement.
❖ Frequency Distribution.
❖ Getting Familiar to Different Statistical Charts.
❖ Which Charts/Tables/Frequency Distribution to choose for which Type of
Variables?

Week 05
 ● Class 13: Measures of Central Tendency & Dispersion
❖ Measures of Central Tendency [Mean, Median, Mode], and Location [Quartile,
Decile, Percentile].
❖ Which Measure is the Best one in which situation?
❖ Measures of Dispersion [Range, Variance, Standard Deviation, Coefficient of
Variation].

● Class 14: Shape Characteristics & Causality

❖ Measures of Shape Characteristics: Skewness, Kurtosis.
❖ Correlation Analysis.
❖ Concepts of Regression Analysis.
❖ What is causality and what is the difference between Correlation and Regression
Analysis?

● Class 15: Inferential Statistics

❖ Normal Distribution & Standard Normal Distribution.
❖ Central Limit Theorem and where this is applied?
❖ Estimators and Estimates.
❖ Confidence Interval and Margin of Error.
❖ Hypothesis Test Concepts: Null & Alternative Hypothesis, Type 1 & Type 2
Error, Test Statistics, Level of Significance & p-value.

Week 06
● Class 16: Parametric Hypothesis Test
❖ Assumptions for Choosing a Parametric Hypothesis Test.
❖ T-tests, F-test, Chi-squared Test.
❖ Executive the Hypothesis Testing in Python and Interpreting the Results.
❖ When to choose which one?

● Class 17: A/B Testing & Non-parametric Hypothesis Tests

❖ What is A/B Testing and why is it so popular among the Data Scientists?
❖ Performing A/B Testing in a Dynamic way in Python.
❖ When to pick Non-parametric Hypothesis Tests?
❖ Mann-Whitney U Test (Wilcoxon Rank-Sum Test), Wilcoxon Signed-Rank Test,
Kruskal-Wallis Test, Kolmogorov-Smirnov Test (for Normality Checking).

*** By completing the Statistics with Python for Data Science module, students
will gain a robust understanding of essential statistical concepts and how to apply
them to real-world data science problems using Python. This module aims to
bridge statistical theory and practical implementation, which is critical for data
analysis and predictive modelling.

Overall Achievement:
By the end of this module, students will have the ability to:

❖ Analyse and interpret data using descriptive and inferential statistical

methods.
❖ Apply Python to execute statistical analyses, enhancing their data science
skills.
❖ Select and implement the appropriate hypothesis tests for various data types
and scenarios.
❖ Understand the role of statistical analysis in data-driven decision-making
processes, preparing them for roles in data science and machine learning.

Data Analysis using NumPy, Pandas

Week 06
Class 18: Data Analysis using NumPy
❖ A brief introduction
❖ Jupyter notebook Installation & exploring.
❖ Google Colab
❖ NumPy arrays, arrange, linspace
❖ Array methods and attributes.
❖ Indexing, slicing
❖ Broadcasting
❖ Boolean masking
❖ Arithmetic Operations
❖ Universal Functions

Week 07
Class 19: Data Analysis using Pandas -1
❖ Pandas Introduction.
❖ Pandas Data Structures - Series
 ❖ Pandas Data Structures – DataFrame
❖ Creating DataFrame
❖ Grab data (column wise)
❖ Grab data (raw wise)
❖ Grabbing an element or a sub-set of the dataframe
❖ Adding new column
❖ Deleting the column
❖ Boolean mask
❖ reset_index(), set_index(), head(), tail(), info(), describe()

Class 20: Data Analysis using Pandas -2

❖ Handling Missing Data
❖ Data Wrangling
❖ Combining, Merging, Joining, Group by
❖ Useful Methods and Operations
❖ unique(), nunique(), value_counts(), count(), sort_value()
❖ Different Ways of Creating DataFrame

Class 21: Exploratory Data Analysis (EDA) using Pandas

Week 08
Class 22: Project on Data Analysis using Pandas

*** NumPy and Pandas are two fundamental Python libraries that form the
cornerstone of data analysis and manipulation. Learning these libraries equips you
with the tools to efficiently work with large datasets, extract meaningful insights,
and make data-driven decisions.

Key Purposes: Data Cleaning and Preparation:

❖ Handling Missing Data: Imputing missing values, removing outliers,

and handling inconsistencies.
❖ Data Formatting: Converting data types, standardising formats, and
cleaning text data.
❖ Data Integration: Merging and joining datasets from various sources.
❖ Efficient Analysis: Structured and organised data facilitates efficient
analysis and visualisation.
 Data Visualization using the most widely used
Python library
Class 23: Data Visualization using Matplotlib
❖ Basic Plotting
❖ Creating multiple plots on the same canvas.
❖ Matplotlib "Object Oriented" approach
❖ Creating a figure and a set of subplots
❖ Saving figures
❖ Decorating the figures

Class 24: Data Visualisation Using Pandas

❖ Histograms
❖ Style Sheets
❖ Area plot
❖ Bar plot
❖ Line plot
❖ Hexbin
❖ Scatter plot
❖ Box Plot
❖ Pie Plot

Week 09

Class 25: Data Visualization using Seaborn

❖ Distribution Plots: Visualising the distribution of a dataset
❖ Joint plot, Pair plot, KDE plot
❖ Plotting with categorical data
❖ Swarm plot and Strip plot
❖ Box plot and Violin plot
❖ Bar plot and Point plot
❖ Matrix Plots, Heatmap

Class 26: Data Visualization using Plotly

❖ Histogram
❖ Bar Chart
❖ Stacked Bar Chart
❖ Bubble Chart
❖ Donut Chart
❖ Scatter Chart
❖ Line Chart
❖ Spider Chart
 ❖ Sunburst Chart
❖ Waterfall Chart

Class 27: Exploratory Data Analysis (EDA) with data Visualisations

***Data visualisation is a powerful technique that helps us understand and

communicate complex information through visual representations. Matplotlib, a
versatile Python library, is a key tool in this process.

Key Purposes of Data Visualization:

Exploratory Data Analysis (EDA):

❖ Identifying Patterns: Spotting trends, seasonality, and outliers.

❖ Understanding Distributions: Visualising the shape and spread of data.
❖ Detecting Anomalies: Identifying unusual data points.

Communicating Insights:

❖ Storytelling: Creating compelling narratives from data.

❖ Simplifying Complex Ideas: Making complex data accessible to a
wider audience.
❖ Highlighting Key Findings: Emphasising important conclusions.

Machine Learning
Data Processing, Transforming and Feature
Engineering

Week 10
Class 28: Introduction to Machine Learning
❖ Introduction to ML - What, Why
❖ Machine Learning Applications
❖ Supervised Learning
❖ Unsupervised Learning
❖ What is a Machine Learning Model?
❖ Training data and Test data
❖ Splitting Data, Train set & Test set
❖ Underfitting and Overfitting
 ❖ KFold Cross Validation
Class 29: Data processing, Transforming, Extractions
❖ Feature Scaling Theory
❖ Feature Scaling - Hands-on
❖ Feature Extractions
❖ Image to Pixel using CV2
❖ Resize, bitwise not and Pixel to Photo
❖ How to prepare a dataset using photos
❖ Linear Discriminant Analysis (LDA)
❖ Bag of words, vectorization
❖ Bag of N-grams
Class 30: Feature Selection, Outlier Detection and Removal
❖ Feature Importance, Feature Selection
❖ Label Encoding
❖ Ordinal Encoding
❖ One Hot Encoding
❖ Outlier Detection and Removal Using IQR
❖ Outlier Detection and Removal using Standard Deviation
❖ Outlier Detection and Removal using Z-Score

***By the end of this module, students will have the ability to

❖ Features Scaling: Scale features as per model requirements (e.g.,

Normalizations Standardization and Absolute ).
❖ Selecting Relevant Features: Identifying the most important features
for a specific task.
❖ Transforming Features: Applying transformations to improve model
performance (e.g., log transformation, square root transformation).
❖ Encoding Categorical Variables: Converting categorical variables into
numerical representations (e.g., one-hot encoding, label encoding).
❖ Outlier Detection & Removal: Finding and removing outliers

Why is Data Processing Important?

❖ Improved Model Performance: Well-prepared data leads to more

accurate and reliable models.
❖ Enhanced Insights: Clean and transformed data allows for deeper
understanding of the underlying patterns.
❖ Efficient Analysis: Structured and organised data facilitates efficient
analysis and visualisation.
 ❖ Better Decision Making: Data-driven decisions based on accurate and
reliable insights.

Supervised Learning (Regression)

Week 11
Class 31: Regression
❖ Simple vs. Multiple Linear Regression Model.
❖ Model Assumptions
❖ Multiple Linear Regression Model in Python using Statsmodels (Statistical
Approach for Estimating the Causality/Dependency).
❖ Interpreting the Model Outputs.
❖ R-squared and Adjusted R-squared.
❖ Hypothesis Tests of Model Coefficients & Goodness of Fit of the Model.
❖ Residual Plots.

Class 32: Regression

❖ Multiple Linear Regression Model in Python using Scikit-learn (Machine
Learning Approach for Prediction Purpose)
❖ Splitting Data into Train-Test-Validation Set
❖ Prediction
❖ Model Evaluation Metrics: R-squared, Mean Squared Error (MSE), Root Mean
Squared Error (RMSE), Mean Absolute Error (MAE), etc.
❖ Feature Importance Plot

Class 33: Regression

❖ Introduction to Regularized Regression Methods
❖ Ridge Regression Concept
❖ LASSO Regression Concept
❖ Elastic Net Regression Concept
❖ Which one to choose?
❖ Fitting a Elastic Net Regression with ElasticNetCV

Supervised Learning (Classifications)

Week 12
Class 34: Class Imbalance, Confusion Matrix & Bias Variance Trade off
❖ Class Imbalance Problem
❖ Class Imbalance Problem Handling
❖ Confusion Matrix (Precision, Recall, f1 score)
 ❖ Bias and Variance, Bias Variance Trade off

Class 35: Binary Logistic Regression Model

❖ Binary Logistic Regression Theory
❖ Confusion Matrix (Precision, Recall, f1 score)
❖ Binary Logistic Regression Algorithm
❖ Binary Logistic Regression Pen & Paper Exercise
❖ Hands-on: Model development using Binary Logistic Regression

Class 36: K Nearest Neighbours

❖ K Nearest Neighbors Theory
❖ K Nearest Neighbors Algorithm
❖ K Nearest Neighbors Pen & Paper Exercise
❖ Model development using K Nearest Neighbors - Hands-on
❖ Hands-on: Handwritten digits recognition using KNN

Week 13
Class 37: Decision Tree
❖ Decision Tree Theory
❖ Decision Tree Algorithm
❖ Decision Tree Pen & Paper Exercise
❖ Hands-on: Model development using Decision Tree

Class 38: Support Vector Machines (SVMs)

❖ Support Vector Machines Theory
❖ Support Vector Machines Algorithm
❖ Support Vector Machines (SVMs)
❖ Hyper-parameter tuning
❖ GridsearchCV
❖ Hands-on: Model development using Support Vector Machines

Class 39: Naïve Bayse Classifications

❖ Naïve Bayse Classifications Theory
❖ Naïve Bayse Classifications Algorithm
❖ Naïve Bayse Classifications Pen & Paper Exercise
❖ Naïve Bayse Classifications
❖ Hands-on: Ham Spam Detection using Naïve Bayse Classification

Week 14
Class 40: Project on Supervised Learning (Classification)
 Unsupervised Learning (Clustering)

Class 41: K Means Clustering

❖ K Means Clustering Theory
❖ K Means Clustering Algorithm
❖ K Means Clustering, Elbow method
❖ K Means Clustering - Hands-on
❖ Hands-on: RFM Based Customer Segmentation using K Means Clustering

Class 42: Principal Component Analysis (PCA)

❖ What is Principal Component Analysis?
❖ PCA vs LDA Analysis
❖ What are Principal Components?
❖ Principal Component Analysis (PCA) Examples
❖ Hands-on: Implement PCA in Python
❖ Predictive Modelling with PCA Components

Week 15
Class 43: Project on Unsupervised Learning (Clustering)

Ensemble Learning
Class-44 Ensemble Learning Models - Random Forest
❖ Ensemble Learning
❖ What is Bagging
❖ What is Bagged Tree
❖ Random Forest
❖ Random Forest Theory
❖ Random Forest Algorithm
❖ Hands-on: Random Forest

Class-45 Ensemble Learning Models – XGBoost

❖ Boosting
❖ XGBoost Theory
❖ XGBoost Algorithm
❖ Hands-on: XGBoost
 Time Series Analysis

Week 16
● Class 46: Introduction to the Basics of Time Series Analysis
❖ Introduction to Time Series Analysis
❖ Moving Averages
❖ Exponential Smoothing
❖ Decomposition

● Class 47: Time Series Analysis with Statistical Modeling

❖ Different Approach to TRAIN-TEST Split for Time Series Forecasting.
❖ ARIMA
❖ SARIMAX
❖ AUTO ARIMA

● Class 48: Time Series Analysis with Prophet library

❖ Forecasting Future Data using Facebook Prophet Library.
❖ Comparing the Results between the Statistical Approach and ML Approach.

Deep Learning (7 Classes + 1 Project)

Week 17
● Class 49: Deep Learning Fundamentals and Model Training
❖ Introduction to deep learning, neural network basics, and model architecture.
❖ Fundamentals of training neural networks, including loss functions and
optimization.

Hands-on: Training a basic neural network on a simple dataset.

● Class 50: Convolutional Neural Networks (CNNs) and Convolution

❖ Introduction to CNNs, the concept of convolution, and layers used in CNNs.
❖ Applications of CNNs in image classification and object detection.

Practical: Building and training a CNN model on an image classification task.

● Class 51: Hyper-parameter Tuning and Optimization

❖ Techniques for tuning deep learning models and managing overfitting.
Hands-on: Optimizing a CNN or OCR model with hyperparameter adjustments.

Week 18
Class 52: Model Evaluation and Validation Techniques
❖ Evaluation metrics and validation techniques for image classification and OCR
models.
❖ Practical: Evaluating the performance of the OCR pipeline and CNN model.

Class 53: Introduction to Image Processing

❖ Overview of Computer Vision and its role in ML and deep learning.

Practical: Preprocessing images including resizing, filtering, and edge detection.

Class 54: Object Detection and Recognition

❖ Introduction to Object detection and principles.

Week 19
Class 55: Real-World Applications
❖ Advanced image processing for region detection.

Practical: Detecting objects in complex images and extracting information.

Class 56: Project Presentation and Discussion

❖ Students work on a project combining learnings
❖ Project presentations, peer feedback, and instructor critique.

MLOps (4 Classes)
Class 57: Introduction to MLOps Concepts
❖ Overview of MLOps and the ML model lifecycle.
❖ Importance of MLOps in production environments and introduction to CI/CD in
ML.
Week 20
Class 58: Building and Managing ML Pipelines

❖ Creating an end-to-end ML pipeline, including preprocessing, training, and

inference.

Practical: Using python to build a preprocessing pipeline within a full ML pipeline.

Class 59: Cloud Deployment of ML Models

❖ Deploying a deep learning model to the cloud and creating an API endpoint.

Hands-on: Deploying a model for real-time access.

Class 60: Monitoring and Maintenance of Deployed Models

❖ Techniques for monitoring model performance and handling model drift.

Practical: Setting up monitoring for the deployed model.

Job Preparation
Extra Week taken by the Support Instructor:

❖ Class 1:
* CV Making & How to Write a Cover Letter?
* Portfolio Building.

❖ Class 2:
* Guidance on Freelancing Career
* Interview Skill Development & How to Effectively Apply for a Position.
* Roadmap for Future Ahead.