You work for a firm that produces cameras that can be used for research studies of animals in the wild.
When placed in the wild, these cameras are used to identify individual animals and groups of animals as they pass in front of the camera.
Researchers use your company's cameras to catalog animal traffic and specific animal counts in geographic areas where these animals are suspected of living.
An example is an identification and counting of wolves in Canada and the far reaches of North America. Using your company's cameras, you and your team of machine learning specialists have been contracted by the Wolf Conservation Center of North America to build a machine learning model to identify and count a specific wolf species in remote areas of the Arctic Circle.
What types of machine learning problem are you trying to solve?
Click on the arrows to vote for the correct answer
A. B. C. D.Answer: B.
Option A is incorrect.
A linear regression is used to model the relationship between a dependent variable and one or more independent variables.
For example: what will the sales in the North American region be when the GDP (Gross Domestic Product) is trending up, and interest rates are trending down.
You are trying to solve a classification problem with images as your inference data.
Option B is correct.
A binary classification is used to classify an observation into one of two categories.
For example: based on the image data, is the animal in the image the wolf species we are looking for or not.
You are trying to solve a binary classification problem: is the animal in the image the species I'm looking for or not? You are looking for a specific species of wolf.
Option C is incorrect.
A multidimensional regression is used to find more than one real number values.
For example: what is the height and width of the animal in the image? You are trying to solve a multiclass classification problem: what type of animal is in the image? You are looking for a specific species of wolf.
Option D is incorrect.
A multiclass classification solves a classification problem where you have more than one class for your answer.
For example: of all the animals identified in a given region, what type of animal is in the image? Of all the types of wolves identified to live in the Arctic Circle, what specific species of wolf is in the image? The problem we're trying to solve is whether this is the specific wolf species we're looking for or not? We are looking for one species class.
Therefore, we should use a binary classification algorithm.
Reference:
Please see the Amazon Machine Learning developer guide titled Formulating the Problem, and the article titled Frame a problem as a machine learning problem or otherwise.
The problem you are trying to solve is to identify and count a specific wolf species in remote areas of the Arctic Circle using machine learning.
Machine learning can be categorized into three main types of problems: regression, classification, and clustering.
Regression models aim to predict continuous numeric values, such as sales or stock prices, based on input variables. Linear regression is a type of regression model that tries to fit a straight line to the data.
Classification models, on the other hand, are used to predict discrete categories or classes. Binary classification is a type of classification model that predicts one of two possible outcomes, such as whether an email is spam or not. Multiclass classification is a type of classification model that predicts one of multiple possible outcomes, such as whether an image contains a cat, dog, or bird.
Finally, clustering models are used to group similar data points together based on their characteristics.
In the case of identifying and counting a specific wolf species, the problem is clearly a classification problem. The goal is to correctly identify whether a particular animal in front of the camera belongs to the specific wolf species or not. Additionally, it is a multiclass classification problem, as there may be multiple species of animals that pass in front of the camera and need to be distinguished from each other.
Therefore, the answer to the question is D. Multiclass classification.