Overview
| Part of |
-
|
| Duration |
5 days
|
| Course Time |
|
| Enquiry |
Please contact ask-iss@nus.edu.sg for more details.
|
The main purpose of this 5-day course is that better data preparation leads to more reliable AI models and better business outcomes.
In many analytics and AI projects, the biggest challenge is not just about building the model, they fail when the data is messy, incomplete, biased, poorly structured, or not machine-readable. Missing values, inconsistent formats, duplicated records, outliers, fragmented sources, and unstructured inputs can weaken model performance, slow down projects, and reduce trust in business insights.
The course focuses on real workplace data challenges. You will work with tabular datasets as well as multimodal data such as text, images, and audio. Through guided exercises and applied workshops, you will practise handling missing values and outliers, merging and reshaping datasets, creating useful features, reducing dimensionality, and converting unstructured data into numerical representations that machine learning models can use.
Participants will learn how to use modern AI tools to accelerate data preparation process while applying the validation, judgement, and quality checks needed for trustworthy analytics and AI outcomes.
By the end of the course, you will be better equipped to turn messy and complex data into high-quality, modelling-ready datasets. This will help you reduce rework, improve model reliability, support better predictive performance, and generate insights that are more useful for business decisions.