DQ 1: Difference between Supervised and Unsupervised Machine Learning

Supervised machine learning is gaining knowledge about a subcategory of machines and artificial intelligence that uses labeled database to train algorithms that helps classifying data outcomes accurately. In Supervised learning, the machine is trained using well-labeled data. One can compare it to the learning process, which takes place in a supervisor or a teacher's presence. The model is fed with data, adjusting its weight until the model fits appropriately. Unsupervised machine learning is machine training that does not need supervision. Instead, the model works for itself in discovering information and deals with unlabeled data. It can be too unpredictable, and it allows one to perform complex processes. Supervised learning allows one to collect data and produce output from previous experiences, helping one optimize performance and solve various real-world problems. On the other hand, unsupervised machine learning helps find all unknown patterns in data helping in finding features useful in categorizing data. Unsupervised learning takes place in real-time, making it easier for learners to get unlabeled data from a computer as it does not need manual intervention compared to labeled data.

The health care sector is one of the early adopters of technological advances of machine learning, playing a key role in many health-related settings, including handling patient data, developing new medical procedures, and treating chronic diseases. Many healthcare uses machine learning to help pathologists make quick and accurate diagnoses of diseases and identify the beneficiaries of the new treatment. Another example is using machine learning to share patient’s medical data privately (Mazlan, Binti Sahabudin, Ramli, Ismail, Mohamad, et al., 2021). A model-trained machine can be used to distribute patient data around the world by inputting patients’ scans, which will be transferred to a centralized server registered as a consensus model and, therefore, get clinically used.

DQ 2: Analytic Techniques Used to Analyze Supervised and Unsupervised Learning

Data analytics are methods used to evaluate data to discover useful information, which is done through cleaning, inspecting, transforming, and modeling data using statistical and analytical tools. Some of the data analysis techniques include regression analysis, which is used to estimate the between different variables and how they might impact the dependent variable to identify patterns and trends that are mostly used in forecasting future trends. For example, I am working in the bank sector, I must examine the relationship between the money spend on social media doing marketing and the sales revenue received. In this case, sales income is the dependent variable I will be interested in boosting and predicting. My independent variable is social media which I will consider whether to increase the money spent on it or not depending on the impact it brings on the dependent variable.

Factor analysis is another technique used to reduce variables to a smaller number of factors. Multiple observable variables correlate because they are all associated with an underlying construct that helps uncover hidden patterns that cannot be easily observed or measured, such as happiness, wealth, or even fitness (Mahdavinejad, Rezvan, Barekatain,  Adibi, Barnaghi, et al., 2018).  For example, working in a bank, one can send questioners to clients with questions relating to how they feel about the company and services offered. Questions like "how do you feel about our company, and would you recommend us to a friend?" will help uncover hidden data.