Steps in UL

UL is a type of ML that allows us to draw inferences from datasets consisting of input data without labeled responses. Unlike SL, where we have a clear target or outcome to predict, UL is more about discovering hidden patterns and structures within data. But how does this process work? Let’s break it down into digestible steps:

Figure 8.1: Steps involved in unsupervised ML (UML)

Note

While the diagram presents a linear flow, in practice, these steps may not always follow a strict linear sequence. Throughout the process, insights gained about the data, such as during evaluation, may inform iterations and refinements in data processing or model selection.

Step 1 – Data collection

Just as with any other ML project, UL begins with data collection. This could be customer data for a retail company, patient data for a healthcare organization, or user behavior data for a tech firm. The key here is to gather as much relevant data...