What is ML?
The machine learning process could be compared to the analogy of
What is teaching a baby/a student?
A REPRESENTATIVE data collection is important because
What is data is what your model is trained on to solve your problem?
Reinforcement learning could be described as
What is a reward-based video game
You are looking for a dataset and go to well-known, diverse data repositories. You go to
What is Kaggle, UCI datasets, Keras, Tensorflow?
Disadvantages of too little data include
What is prone to bias & generalized trends?
The most critical part of the machine learning process is
What is data collection?
The sets data is split into is called
What is training, validation, & testing?
Supervised learning is used _ data and could be described as
What is labeled data and someone giving you the answers to identify?
Choose your favorite reward-based video game. Imagine you have started playing it and haven't been successful in your strategy to attain more rewards. What do you do?
What is [depends on the validity and soundness of your answer]?
What is k-fold, leave-one-out, stratified, time-series, random, etc?
AI originated from
What is the calculator?
What is raw, diverse, recent, less noisy, set dimensional data?
Unsupervised learning is used for _ data and is described as
What is unlabeled data and pattern analysis?
You are taking care of a toddler. You are trying to teach the toddler some shapes with the books you have (the books include labeled data), however, the baby just seems to not learn anything and cry. For effective learning, you decide to
What is changing the books (to something more interactive) and/or giving the toddler a treat every time it learns a shape right?
Advantages of too much data include
What is improved accuracy metrics, higher efficiency for deep learning models, reduced bias (if REPRESENTATIVE), and better EDA results?
The machine learning process can be described as (Hint: 5 to 6 steps, think data)
What is data collection, data preparation, EDA (data analysis), machine learning model, and visualization?
The types of data include
What is labeled, unlabeled, numeric, categorical, and ordinal?
Subcategories of supervised learning include
What is classification and regression?
You are a realtor in the area and want to use AI to your advantage to better understand current trends and projected house prices. The variables you would account for are
What is city, cost of living(#), quality of life(#), current house prices, population, rate of population growth, etc.?
Disadvantages of too much data include
What is computationally expensive, increased training times, increased complexity, and diminishing results?
Machine learning can be categorized into the 5 subcategories of
What is supervised, unsupervised, reinforcement, deep learning, and deep reinforcement learning?
The process(s) to choose a REPRESENTATIVE dataset includes
What is EDA, PCA, &/or LDA?
An example that could describe supervised learning is
What is any type of classification or regression?
You work in the biomedical industry and want to make an autonomous robots that organizes medicine based on time it needs to be taken and dosage. You have a container you can design, prescription, the medicines, a camera, and any robot material. You follow the machine learning process to make this robot and end up with
What is [we'll judge based on your answer]?
Advantages of too little data include
What is reduced storage requirements, faster training times, increased focus on data quality, & better suited for most quantum tasks?