Data Science syllabus

Basic Concepts of Data: Types of data: Structured, Semi-structured, Unstructured, Scales of measurement: Nominal, Ordinal, Interval, Ratio, Data formats: CSV, JSON, XML, SQL tables, Data quality dimensions: Accuracy, Completeness, Consistency, Timeliness, Difference Between Data, Information, Knowledge, wisdom. Definition of Data Science, The Data Science Process/Lifecycle (e.g., CRISP-DM, Team Data Science Process (TDSP), SEMMA Methodology), Applications of Data Science in various domains (e.g., Healthcare, Finance, Retail), Distinction between Data Science, Machine Learning, and Artificial Intelligence, Roles in Data Science (e.g., Data Analyst, Data Scientist, Machine Learning Engineer), Ethical and Privacy Issues in Data Science.

Case Study: A telecom business wants to proactively identify customers who are likely to quit because of the high rate of customer turnover. The organization wants to predict loss of customers using data science techniques and take proactive steps to keep at-risk clients. Determine the steps that, as data scientists, we must adhere to in order to resolve this issue.

Importance of mathematical statistics in Data Science, Role of mathematical statistics in building data models, mathematical statistics foundations needed for machine learning algorithms. Linear Algebra for Data Science: Vectors and Matrices, Vector operations (dot product, cross product) and its application in data science, Matrix operations (addition, multiplication, inverse, determinant), Eigenvalues and Eigenvectors and its uses in dimensionality reduction, Probability Concepts: Random variables and probability distributions, Bayes’ Theorem and conditional probability, Expected value and variance. Statistical Methods: Measures of central tendency (Mean, Median, Mode), Measures of dispersion (Variance, Standard Deviation) and its importance in data preprocessing ,Correlation and covariance and its importance in feature selection in modeling.

Case Study : In a dataset, you have a feature matrix and output vector .Elaborate, How does matrix multiplication help in solving linear regression using the Normal Equation.

Machine Learning: Types (Supervised, unsupervised, Reinforcement), Supervised Algorithms (Linear Regression, Logistic Regression, Decision Trees, SVM). Metrics: Confusion Matrix, Accuracy, Precision-Recall, ROC-AUC, F1 Score, Mean Squared Error (MSE),R-squared. NumPy: Arrays, Mathematical Operations, Linear Algebra, Random Number Generation, Broadcasting. Pandas: Data Frames, Cleaning, Transformation, Filtering, Grouping, Import/Export, Missing Data Handling. Matplotlib: Basic & Advanced Plots, Customization, Subplots, Interactive Features. Seaborn: Distribution, Categorical, Relationship, Regression Plots, Multi-Plot Grids, Themes.

Case study: Write a case study on E-commerce Data. Key Components of the Analytical Plan are: Problem statement and objectives, Data Collection and Preparation, Data Analysis and Model Development, Analysis Outcomes and Takeaways

Data Preprocessing, Data Cleaning: Handling missing data, outliers. - Data Transformation Techniques: Feature Extraction and Selection (PCA), encoding. - Exploratory Data Analysis (EDA): Univariate, bivariate, multivariate analysis. Outliers: (Z-Score Method, Interquartile Range) Data Visualization: Overview of Data Visualization, Need of Data Visualization, Shapes of data, input for data visualization, Types of Data Visualization: Cognitive and perceptual, Practicing good ethics in Data Visualization, Principles of visual perception, Data Visualization Tool (Tableau, Power BI)

Case study : 1.

Case Study on Exploratory Data Analysis (EDA) and Visualizations 2. Create simple plot to visualize a distribution of variables using python

ETL, ETL Challenges, ETL in Data Preprocessing for AutoML, AutoML: AutoML Workflow, AutoML Libraries: Auto-sklearn, H2O.ai, and TPOT, Benefits and Limitations of AutoML, AutoML vs Traditional Machine Learning, Pandas for data preprocessing for AutoML, data normalization and standardization for AutoML, Open-source Tools and Environments: Python, R, Jupyter, Git, VS Code and its roles in data science. Introduction to Big Data: Characteristics, Architecture and Ecosystem, Tools and Technologies, Applications and Challenges of Big Data in Data Science. AI-powered Data Cleaning: Using AI models to clean and structure raw data before analysis.

Case study: A retail company wants to predict its future sales revenue based on various factors such as product pricing, marketing spends, seasonal trends, and customer demographics. The company aims to use regression and classification models to analyze historical sales data, identify trends, and make data-driven decisions to enhance profitability and optimize resource allocation. Justify how does the choice between regression and classification models affect sales predictions?