Machine Learning

Question 1

What is Machine Learning?

Answer 1

• subset of AI
• focused on algorithms that learn and make predictions based on data
• systems improve their performance over time without hand-coded instructions

Question 2

What are the types of Machine Learning?

Answer 2

• Supervised: uses labeled datasets eg cancerous vs healthy cells, applied in spam filters, recommendation engines
• Unsupervised: data lacks explicit labels or guidance, eg clustering data with shared attributes
• Reinforcement: interact with an environment, learns from feedback (rewards or penalties)

Question 3

How do AI text detectors differentiate between AI and human written text?

Answer 3

• Perplexity: how surprising or unpredictable the text is; human writing more varied and creative with high perplexity
• Burstiness: variability in sentence length and structure; ai writing more uniform and consistent with low burstiness

Question 4

Define Overfitting in Machine Learning.

Answer 4

• model performs well on training data
• but fails to generalize to new, unseen data.

Question 5

What is a Confusion Matrix?

Answer 5

• table that evaluates performance of a machine learning model
• assess accuracy and error rates.