Artificial Intelligence 26/27 – BD107

Module Question 3

1. What is the key difference between supervised and unsupervised learning? Provide an example for each.
2. What is “overfitting” in a machine learning model, and why is it a problem?
3. Describe the three main components of reinforcement learning: agent, environment, and reward.
4. How does a training dataset differ from a testing dataset, and why is this separation important?

Answer:

1. The main difference between supervised and unsupervised learning is whether the data used to train the model has labels or not. In supervised learning, the model is trained using data that already has known answers or labels. The model learns from these examples and then uses what it has learned to make predictions on new data. For example, a supervised learning model can be trained using pictures of cats and dogs that are already labelled as “cat” or “dog.” After learning from these examples, the model can predict whether a new picture is a cat or a dog.In contrast, unsupervised learning uses data without predefined labels. The model tries to find patterns, groups, or relationships within the data on its own. For example, a company can use unsupervised learning to group customers based on their shopping habits without previously defining the customer groups.In simple terms, supervised learning learns from labelled data, while unsupervised learning looks for patterns in data without labelled answers.
2. Overfitting happens when a machine learning model learns the training data too closely, including details or patterns that are not actually useful. As a result, the model may perform very well on the training data but have difficulty making accurate predictions when it receives new data. For example, if a model is trained to recognize cats and dogs, it might memorize specific features of the training pictures instead of learning the general characteristics of cats and dogs. When it receives a new picture, its performance may decrease. Overfitting can be reduced in several ways, such as using more training data, simplifying the model, applying regularization, and using techniques such as cross-validation. These methods help the model learn general patterns instead of simply memorizing the training data.
3. The purpose of splitting data into training and testing sets is to evaluate how well a machine learning model can perform on new data. The training set is used to teach the model and help it learn patterns from the data, while the testing set is used after training to check whether the model can make accurate predictions on data it has not seen before. This is important because a model may perform very well on the training data simply because it has learned the data too closely. By using separate testing data, we can get a better idea of how well the model can generalize to new situations. For example, if we have 1,000 customer records, we could use part of the data to train the model and keep the remaining data for testing. The testing data acts like a new situation for the model and helps us evaluate its actual performance.
4. Classification and regression are both types of supervised learning, but they are used for different kinds of predictions. Classification is used when the output is a category or class. For example, a model can predict whether an email is spam or not spam. Regression, on the other hand, is used when the output is a numerical value. For example, a machine learning model can predict the price of a house based on factors such as its size, location, and number of rooms. In simple terms, classification predicts a category, while regression predicts a numerical value.

Module 3 – Explanatory Video click here

Status: 100% telah terpenuhi

Keterangan: Module question dan penjelasan jawabannya sudah tersedia dalam bentuk vidio animasi.