What I learned
Machine learning can be achieved through supervised learning, unsupervised learning, and reinforcement learning. There is also semi-supervised learning which is the intermediate between supervised and unsupervised learning, as the name suggests. Supervised learning takes in labeled data and learns which label represents which feature. Unsupervised learning takes in unlabeled data to identify a previously hidden pattern. Finally, reinforcement learning is based on the principle of feedback and improves its performance over trials.
Supervised learning can have two types of applications: regression and classification. Regression predicts continuous values that do not have a specific constraint or limit. Classification predicts discrete values.
Unsupervised learning also has two types of application: clustering and association. Clustering involves detecting similarities in the unlabeled data, in which the clusters are then categorized. Association is when the machine finds a relationship between the data and underlying patterns.
There are numerous uses of machine learning that range from patient diagnosis and speech recognition to fraud detection. While I was researching, I recalled a project I made in my first CS course where we programmed an AI connect four. The AI sought out the best move from the current board and rejected the moves that would make the AI lose. This seems like a version of reinforcement learning.
Log in or sign up for Devpost to join the conversation.