Tuesday November 19, 2019
Home Lead Story Machine Learn...

Machine Learning and AI: The Puzzle is Not Solved Yet

Fifty years ago, a chess-playing programme was considered a form of AI

0
//
Rana el Kaliouby, CEO of the Boston-based artificial intelligence firm Affectiva, is pictured in Boston, April 23, 2018. Affectiva builds face-scanning technology for detecting emotions, but its founders decline business opportunities that involve spying on people.
Rana el Kaliouby, CEO of the Boston-based artificial intelligence firm Affectiva, is pictured in Boston, April 23, 2018. Affectiva builds face-scanning technology for detecting emotions, but its founders decline business opportunities that involve spying on people. VOA

By Nishant Arora

The most buzzed-about disruptive technologies that are changing business landscapes today are Machine Learning (ML) and Artificial Intelligence (AI). Almost all of us have heard or read about them but do we actually know what the fuss is all about?

The enterprises are trying to harness the explosion of digital data and computational power with advanced algorithms to enable collaborative and natural interactions between people and machines.

However, there’s still a lot of confusion within the public and the media regarding what is ML and AI.

People prefer to write AL and ML technologies — and not ML and AI — and the argument goes that the former syncs well with the human mind.

Both the terms are often being used as synonyms and in some cases as discrete, parallel advancements.

In reality, ML is to AI what neurons are to human brain. Let us start with ML.

According to Roberto Iriondo, Editor of Machine Learning Department at Carnegie Mellon University in Pennsylvania, ML is a branch of AI.

As coined by computer scientist and machine learning pioneer Tom M. Mitchell, “ML is the study of computer algorithms that allow computer programmes to automatically improve through experience”.

For instance, if you provide an ML model with songs that you enjoy, along with audio statistics (dance-ability, instrumentality, tempo or genre), it will be able to automate and generate a system to suggest you music that you’ll enjoy in the future, similarly as to what Netflix, Spotify and other companies do.

“In a simple example, if you load an ML programme with a considerable large data-set of X-ray pictures along their description (symptoms etc), it will have the capacity to assist (or perhaps automatise) the data analysis of X-ray pictures later on,” said Iriondo.

The ML model will look at each one of the pictures in the data-set, and find common patterns in pictures that have been labelled with comparable indications.

Shanghai,
Rana el Kaliouby, CEO of the Boston-based artificial intelligence firm Affectiva, demonstrates the company’s facial recognition technology, in Boston, April 23, 2018. VOA

AI, on the other hand, is exceptionally wide in scope and is a system in itself and not just independent data models.

In simpler terms, AI means creating computers that behave in the way humans do.

However, according to Theo van Kraay, Cloud Solution Architect (Advanced Analytics & AI), Customer Success Unit at Microsoft, any attempt to try to define AI is somewhat futile, since we would first have to properly define “intelligence”, a word which conjures a wide variety of connotations.

“Firstly, it is interesting and important to note that the technical difference between what used to be referred to as AI over 20 years ago and traditional computer systems, is close to zero,” says van Kraay.

What AI systems today are doing reflects an important characteristic of human beings which separates us from traditional computer systems – human beings are prediction machines.

Many AI systems today, like human beings, are mostly sophisticated prediction machines.

“The more sophisticated the machine, the more it is able to make accurate predictions based on a complex array of data used to train various (ML) models, and the most sophisticated AI systems of all are able to continually learn from faulty assertions in order to improve the accuracy of their predictions, thus exhibiting something approximating human intelligence,” van Kraay said.

Most ML algorithms are trained on static data sets to produce predictive models, so ML algorithms only facilitate part of the dynamic in the definition of AI.

Fifty years ago, a chess-playing programme was considered a form of AI.

Also Read- ISRO to Launch Radar Imaging Satellite Soon

Today, a chess game would be considered dull and antiquated, due to the fact that it can be found on almost every computer.

“AI today is symbolised with human-AI interaction gadgets like Google Home, Apple Siri and Amazon Alexa or ML-powered video prediction systems that power Netflix, Amazon and YouTube,” says Iriondo.

In contrast to ML, AI is a moving target and its definition changes as its related technological advancements turn out to be further developed.

“Possibly, within a few decades, today’s innovative AI advancements will be considered as dull as flip-phones are to us right now,” quips Iriondo. (IANS)

Next Story

This AI Tool Can Predict Mortality Of Heart Failure Patients

Researchers develop a tool that can predict mortality of heart failure patients

0
heart
This artificial intelligence (AI) tool can predict life expectancy in heart failure patients. Pixabay

Researchers have developed an artificial intelligence (AI) tool to predict life expectancy in heart failure patients.

The machine learning algorithm based on de-identified electronic health, records data of 5,822 hospitalised or ambulatory patients with heart failure at UC San Diego Health in the US.

“We wanted to develop a tool that predicted life expectancy in heart failure patients, there are apps where algorithms are finding out all kinds of things, like products you want to purchase,” said Avi Yagil, Professor at University of California.

“We needed a similar tool to make medical decisions. Predicting mortality is important in patients with heart failure. Current strategies for predicting risk, however, are only modestly successful and can be subjective,” Yagil added.

From this model, a risk score was derived that determined low and high risk of death by identifying eight readily available variables collected for the majority of patients with heart failure:Diastolic blood pressure, Creatinine, Blood urea nitrogen, White blood cell count, Platelets, Albumin and Red blood cell distribution.

Yagil said the newly developed model was able to accurately predict life expectancy 88 per cent of the time and performed substantially better than other popular published models.

“This tool gives us insight, for example, on the probability that a given patient will die from heart failure in the next three months or a year,” said researcher Eric Adler.

Heart failure patients
The mortality of a heart failure patient can be predicted. Pixabay

“This is incredibly valuable. It allows us to make informed decisions based on a proven methodology and not have to look into a crystal ball,” he added.

The tool was additionally tested using de-identified patient data from the University of California San Francisco and a data base derived from 11 European medical centers.

“It was successful in those cohorts as well,” said Yagil.

“Being able to repurpose our findings in independent populations is of utmost importance, thus validating our methodology and its results,” Yagil added.

Also Read-Therapy Dogs Befriend Patients in ICU

Researchers said the partnership between physicists and cardiologists was critical to developing a reliable tool and extensive knowledge and experiences from both sides proved synergetic.

The study was published in the European Journal of Heart Failure. (IANS)