AI for Girls

IN COLLABORATION WITH NGEE ANN POLYTECHNIC

Our Empower: AI for Girls course has successfully been running every term at Ngee Ann Polytechnic for the past few years, giving girls from all polytechnics the opportunity to learn basic skills in AI.

Learning Model

  • Self-directed learning for 9 weeks + 10th-week Capstone project presentation.

  • A fixed set of 5 modules to go through at the same starting point. Students are free to advance as quickly as they want.

  • Optional modules at the back for students to go through if they have completed the fixed set of modules

Learning Outcomes

Encourage the girls to learn and grow as far as they can go in the area of Data Analytics, namely to

● Understand a business problem

● Know how to handle and manipulate data

● Perform exploratory data analysis

● Be able to interpret and perform data visualization

● Know fundamental Machine Learning models to solve different problems

● Appreciate the opportunities and challenges of applying AI in real life.

Curriculum

MODULES

  • Introduction to Python

  • Intermediate Python for Data Science

  • Pandas Foundation

  • Introduction to Data Visualization with Python

  • Supervised Learning with scikit-learn

  • (Optional) Unsupervised Learning in Python

  • Capstone Project

Week 1-4

    • Create variables and understand data types Store, access, manipulate data in lists.

    • Use functions, methods, and packages to reduce amount of code.

    • Learn to work with Numpy array and use it to efficiently do data science

    • Build various plots and customize them to make them visually appealing and interpretable using Matplotlib.

    • Create, manipulate, and access data from Dictionary and Pandas.

    • Learn comparison operators, combine them and use boolean outcomes in control structures.

    • Learn loops to iterate over all kinds of data structures.

    • Build Pandas DataFrame, how to import and inspect datasets using Pandas.

    • Ingest, inspect, and explore your data visually and quantitatively.

    • Manipulate and visualize time series data with tools like upsampling, downsampling and interpolation

Week 5-10

    • Customize plots through methods like overlaying, making splots, and controlling axes.

    • Use, present and orientate grids to represent two- variable functions.

    • Use seaborn to compute and visualize linear regressions and univariate and multivariate distributions.

    • Analyze time series and images.

    • Introduction to classification problems uses regression to solve a problem that requires a continuous outcome.

    • Evaluate a model's performance and the metrics to use to gauge how good it is.

    • Learn about pipelines, transformers, estimators and pre-processing techniques.

    • Create variables and understand data types Store, access, and manipulate data in lists.

    • Use functions, methods, and packages to reduce the amount of code.

    • Learn to work with Numpy array and use it to efficiently do data science

    • Unsupervised Learning in python (optional)

    • Students work in teams of 3 -4 (mixed sch) on group projects, applying the techniques they have learnt.

    • Allow students to provide an in-depth data analysis with visualization or apply Machine Learning models.