AN INSIGHT ON MACHINE LEARNING

Introduction towards machine learning

Machine Learning in simple words Machine learning is a method of data analysis that automates analytical model building. It is a branch of artificial intelligence based on the idea that machines should be able to learn and adapt through experience.

It is a data analytics technique that teaches computers to do what comes naturally to humans and animals: learn from experience. Machine learning algorithms use computational methods to “learn” information directly from data without relying on a predetermined equation as a model. The algorithms adaptively improve their performance as the number of samples available for learning increases.

The process of learning begins with observations or data, such as examples, direct experience, or instruction, in order to look for patterns in data and make better decisions in the future based on the examples that we provide. The primary aim is to allow the computers learn automatically without human intervention or assistance and adjust actions accordingly.

The origin of it

1950 — Alan Turing creates the “Turing Test” to determine if a computer has real intelligence. To pass the test, a computer must be able to fool a human into believing it is also human.

1952 — Arthur Samuel wrote the first computer learning program. The program was the game of checkers, and the IBM computer improved at the game the more it played, studying which moves made up winning strategies and incorporating those moves into its program.

1957 — Frank Rosenblatt designed the first neural network for computers (the perceptron), which simulate the thought processes of the human brain.

1967 — The “nearest neighbor” algorithm was written, allowing computers to begin using very basic pattern recognition. This could be used to map a route for traveling salesmen, starting at a random city but ensuring they visit all cities during a short tour.

Today, machine learning algorithms enable computers to communicate with humans, autonomously drive cars, write and publish sport match reports, and find terrorist suspects.

Why is it a trending technology

The modern challenges are “high-dimensional” in nature. With rich data sources, it is important to build models that solve problems in high-dimensional space. Through it, the models can be integrated into working software. It supports the kinds of products that are being demanded by the industry.

Also, Google Trends that tracks the popularity of search terms, suggests that searches for machine learning are about to out-pace the searches for artificial intelligence. Machine learning is moving beyond the textbooks and is creating a disruption which will revolutionize the future.

In 2014 Machine Learning and AI experts thought it would take at least 10 years before a machine could beat the world’s best player at the board game Go. But Googles DeepMind proved them wrong. They showed that even in such a complex game as Go machines could learn which moves to consider. There are a lot more of advances in the field of machines playing games like the Dota Bot from the OpenAI Team.

Machine Learning is going to have huge effects on the economy and living in general. Entire work tasks and industries can be automated and the job market will be changed forever.

The internal functioning

The main focus is to provide algorithms which can be trained to perform a task. It is closely related to the field of computational statistics as well as mathematical optimization. It contains multiple methods like Supervised Learning, Unsupervised Learning, Semi-supervised Learning, and Reinforcement Learning which each has their own use cases and algorithms.

Supervised Learning maps an input to an output based on example input-output pairs also known as labelled data. Supervised Learning has two sub classes: Classification & Regression. Unsupervised Learning deals with unlabeled data (data that doesn’t have a label). Unsupervised Learning is mostly used for finding relationships in datasets, reducing dimensionality or identifying anomalies. Semi-supervised learning is a mixture of supervised and unsupervised learning. It typically works with a small amount of labelled data and a large amount of unlabeled data. Reinforcement Learning deals with how an agent takes actions in an environment to maximize a reward. 

Choosing the right algorithm can seem overwhelming—there are dozens of supervised and unsupervised machine learning algorithms, and each takes a different approach to learning.

There is no best method or one size fits all. Finding the right algorithm is partly just trial and error—even highly experienced data scientists can’t tell whether an algorithm will work without trying it out. But algorithm selection also depends on the size and type of data you’re working with, the insights you want to get from the data, and how those insights will be used.

• Choose supervised learning: if you need to train a model to make a prediction--for example, the future value of a continuous variable, such as temperature or a stock price, or a classification—for example, identify makes of cars from webcam video footage.
• Choose unsupervised learning: if you need to explore your data and want to train a model to find a good internal representation, such as splitting data up into clusters.

Understanding the working in brief

Machine learning algorithms learn, but it’s often hard to find a precise meaning for the term learning because different ways exist to extract information from data, depending on how the machine learning algorithm is built. Generally, the learning process requires huge amounts of data that provides an expected response given particular inputs. Each input/response pair represents an example and more examples make it easier for the algorithm to learn. That’s because each input/response pair fits within a line, cluster, or other statistical representation that defines a problem domain.

Machine learning is the act of optimizing a model, which is a mathematical, summarized representation of data itself, such that it can predict or otherwise determine an appropriate response even when it receives input that it hasn’t seen before. The more accurately the model can come up with correct responses, the better the model has learned from the data inputs provided. An algorithm fits the model to the data, and this fitting process is training.

How it is changing the world

Managing data can be crucial in the field like education. Smart classrooms have been developed into expanding the database of resources. Digital system can record every individual performance and can provide an accurately customized report of their specific need. With classroom strength increasing day by day this kind of technology help will be a breakthrough in education. Machine learning integrated alarm system and surveillance cameras are very popular today. Machine learning uses facial recognition technology to build a catalog of frequent visitors at home and recognizes unusual visitors.

It is being increasingly used in healthcare for faster patient diagnosis. Machine learning programs can predict health problems based on age, socioeconomic status, and genetic history which helps prevent illness. Hospitals are currently using it for accurately detecting tumors in radiology scans and detecting cancer. Computers can use large data sets and an algorithm to classify the images from scans. Machine learning algorithm has been written that can detect cancer more accurately than the best pathologist, freeing doctors up to make the treatment decision more accurately and quickly. Fully automatic driver less cars are the most prominent display of this technology. Driver less cars can differentiate between trees and pedestrians, fields and roads, and many road signals, which has opened a lot of opportunities in goods delivery and personal transportation.

Machine learning enables an analysis of a massive quantity of data and can provide a faster and more accurate result that can help in identifying profitable opportunities and dangerous risks.

Boundaries of machine learning

 Error diagnosis and correction: One notable limitation of machine learning is its susceptibility to errors. Brynjolfsson and McAfee said that the actual problem with this inevitable fact is that when they do make errors, diagnosing and correcting them can be difficult because it will require going through the underlying complexities of the algorithms and associated processes.

 Time constraints in learning: It is impossible to make immediate accurate predictions with a machine learning system. Remember that it learns through historical data. The bigger the data and the longer it is exposed to these data, the better it will perform. For example, using a system to play games and beat human opponents would require feeding the system with historical data and continuously exposing it to newly acquired data to make better predictions or decisions.

Problems with verification: Another limitation of machine learning is the lack of variability. Brynjolfsson and McAfee said that machine learning deals with statistical truths rather than literal truths. In situations that are not included in the historical data, it will be difficult to prove with complete certainty that the predictions made by a machine learning system is suitable in all scenarios.

 Limitations of predictions :Brynjolfsson and McAfee reminded that unlike humans, computers are not good storytellers. Machine learning systems cannot always provide rational reasons for a particular prediction or decision. They are also limited to answering questions rather than posing them. In addition, these systems does not understand context. Depending on the provided data used for training, machine learning is also prone to hidden and unintentional biases. Human input is still important to better evaluate the outputs of these systems.