Algorithms + Society

Artificial Intelligence (AI)

​

​Artificial intelligence describes the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.

 

 

Machine Learning (ML)

​

Machine learning is a type of AI that focuses on dealing with large amounts of data to create insights by running the data through some sort of algorithm. Machine learning has two types: supervised and unsupervised.

​

Supervised

In supervised learning, the program relies on human intervention. A person provides the base knowledge for the program so it can then understand what to do and how to improve. For example, in classification, a person provides a program with correctly labeled data, and then the program tries to correctly label new data based on that previous data.​

 

Unsupervised

In unsupervised learning, the program doesn’t require this human validation component. It performs the research, identifies new knowledge, and memorizes it all on its own. For example, in clustering, a program receives a large amount of data and tries to section it off into groups of similar seeming data.

​

​

Big Data

​

Big data is a term that describes the large volume of data that exists - both structured and unstructured. It’s not as important that there’s a lot of data as it is how companies use this mass of data to generate insights and business strategies.

 

 

Binary Variables

​

Binary variables are variables that only have two unique values, for example, a variable “Life Status” can contain only two values like “Dead” or “Alive”.

 

 

Classification

​

Classification is a type of supervised machine learning where the output is a category, such as “Cat” or “Dog”. For example, if you had a set of pictures of cats and dogs in front of you, you would try to label each picture as either “Cat” if there was a cat in the picture, or “Dog” if there was a dog in the picture. There are certain types of algorithms used for classification, such as logistic regression or decision trees.

 

 

Clustering

​

Clustering is an unsupervised machine learning method used to discover groupings that exist in the data. For example, a company might group customers based on their purchasing behavior and then recommend certain products to those customers. There are certain types of algorithms used for classification, such as K-means.

 

 

Computer Vision (CV)

​

CV is a field of computer science that works to allow computers to visualize and identify images and videos just like human vision does. One example of CV is pedestrian detection in self-driving cars.

 

 

Accuracy

​

• Accuracy is a metric used to determine how good a machine learning model is, and is represented by comparing the number of correctly predicted classes to the total predicted classes.

  • True-positive = predicted good, is actually good.

  • True-negative = predicted bad, is actually bad.

  • False-positive = predicted good, is actually bad.

  • False-negative = predicted bad, is actually good.

​

​

Precision + Recall

​

Precision and recall are both used to measure the correctness of a model.​

• Precision can be measured as the total actual positive cases​.​​

• Recall can be measured as to how many of the positive predictions were correct.

​

​

Bias

​

Explicit

• Explicit bias means that the person is very clear about his or her feelings and attitudes, and their related behaviors are intentional.

​​

Implicit

• Explicit bias means that the bias exists outside of the person’s awareness, and can be in direct contradiction to the person’s expressed beliefs and values. This can be dangerous because the bias exists outside of the person’s awareness, but can still affect their behavior.