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 to live.
An example is the 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 determine if the animals captured by the camera are a specific species of wolf or not. What type 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 machine learning problem that the team of machine learning specialists are trying to solve is multiclass classification.
Multiclass classification is a type of machine learning problem where the goal is to classify input data into one of three or more possible classes. In this case, the input data is images captured by the company's cameras, and the classes are different species of wolves. The machine learning model needs to be able to identify if the animal captured by the camera is a specific species of wolf or not.
Linear regression is a type of machine learning problem where the goal is to predict a continuous numerical value, such as the price of a house or the amount of rainfall in a certain area. Binary classification is a type of machine learning problem where the goal is to classify input data into one of two possible classes. Multidimensional regression is a type of machine learning problem where the goal is to predict multiple continuous numerical values simultaneously.
Therefore, in this scenario, the goal is to classify the input data (images captured by the company's cameras) into one of several possible classes (different species of wolves), making it a multiclass classification problem.