Lab Notebook Week 16

Crowdsourcing is a method of obtaining information, ideas, services, or content by soliciting contributions from a large group of people, particularly from an online community, rather than from traditional employees or suppliers. This approach leverages the collective intelligence and skills of a vast, diverse crowd to solve problems, generate ideas, and complete tasks.
A public dataset is like a treasure trove of information that’s open for anyone to explore and use. Think of it as a collection of data that’s made available to the public by organizations, researchers, or governments. These datasets cover a wide range of topics— from social and economic trends to scientific research and more. The idea is to encourage collaboration, innovation, and transparency by allowing people to analyze and derive insights from the data. It’s like sharing knowledge on a grand scale!
distributed computing: donates computing power to help other app functions.
the idea that a large group of people can collectively contribute more diverse and substantial input than an individual or small team.
this is usually cheaper and takes less resources

Computing bias is a bias that exists within algorithms that unfairly discriminate against certain individuals or groups, creating unfair outcomes
Data bias: when the data set does not reflect the real world values
Human bias: People who make the programs may bring in their own biases whether consciously or unconsciously
Algorithmic bias reflects the data bias and human bias and often amplifies it. However, bias can arise from the structure and decision making of the program itself
Explicit data is information you give an app or site
Implicit data is the information it collects based on your activity
input correction: Imagine you have a machine learning model that identifies objects in images. Input correction involves making adjustments to the images used to test the model after it has already been trained.
Classifier Correction: This step focuses on fine-tuning the algorithm or model after training to reduce any biases or discrimination it might have inadvertently learned.
Output Correction: After the model makes predictions, output correction involves modifying those predictions to eliminate any biases or unwanted discrimination.

An decidable problem is a decision problem for which an algorithm can be written to produce a correct number for all inputs
An undecidable problem is one for which no algorith can be constructed that is always capable of providing a correct yes-or-no answer.
The essence of the Halting Problem revolves around creating an algorithm that can accurately determine, for any program and input, whether that program will halt or continue running forever. Alan Turing proved that such an algorithm cannot exist.

Lab Notebook Week 16 • 7 min read