Excercise-Hands-On-Machine-Learning-with-Scikit-Learn-and-TensorFlow

Solutions to the exercise questions of the book "Hands-On Machine Learning with Scikit-Learn and TensorFlow"

Chapter-1 Excercise solutions

---

Q1 How would you define Machine Learning?
Sol- Machine learning is all about making a machine learn better from the data rather than expliciting giving instruction or code rules.

Q2 Can you name four types of problems where it shines?
Sol-

1. When there is a lot of rules or hand tuning is needed.
2. Complex problems which do not have any solutions
3. Highly varying systems.
4. Getting insights from large amounts of data

Q3 What is a labeled training set?
Sol- Labeled training set is a set of data points for which the corresponding class information is provided. Systems can use this data to learn the given classes and classify on new data. This type of data is used in supervised learning.

Q4 What are the two most common supervised tasks?
Sol-

1. Classification
2. Regression

Q5 Can you name four common unsupervised tasks?
Sol-

1. Clustering
2. Dimensionality reduction
3. Embeddings
4. Anomaly detection

Q6 What type of Machine Learning algorithm would you use to allow a robot to walk in various unknown terrains?
Sol- Reinforcement Learning

Q7 What type of algorithm would you use to segment your customers into multiple groups?
Sol- Hirarchichel Clustering