Data Analysis for Text, Timeseries and Images with Python

Many of today’s most valuable datasets extend beyond traditional tables. Text documents, digital images, and time-series data are examples of multimodal data that contain rich information requiring specialised analytical techniques. This course equips participants with the skills to process, transform, and analyze these complex data types using Python, expanding their analytical toolkit and enabling them to extract meaningful insights from a wider range of real-world data sources.

While Data Analysis for Text, Time Series and Images with Python can be taken as a standalone course, it is the third of four Microcredentials leading to the Certificate of Open Studies (COS) in Applied Data Analytics and Visualization.

Upon successful completion of this course, participants will be able to:

• Process unstructured text using cleaning techniques, regular expressions, and tokenization
• Vectorize text using TF-IDF weighting to extract meaningful features for downstream analysis
• Manipulate digital images as multidimensional numerical arrays using NumPy
• Analyze image properties such as format, compression, bit depth, and colour channels.
• Transform time-indexed data using Pandas operations – including shifting, differencing, resampling, and rolling windows
• Diagnose temporal patterns (such as trends, seasonality, and cycles) by using visualisation tools and autocorrelation functions
• Assess time-series stationarity using statistical tests and visual diagnostics
• Apply structural decompositions and transformations to prepare time-series data for further analysis

Stackable into Certificate of Open Studies (COS) in Applied Data Analytics and Visualisation

Nongraded

Professionals who already work with structured datasets and want to broaden their analytical toolkit to include text, image, and time-series data. The course is intended for data analysts, researchers, engineers, software developers, and other professionals looking to apply Python to a wider range of real-world data sources.