Basics of Artificial Intelligence, Machine Learning, and Deep Learning
Artificial Intelligence is buzzing in every tech discussion and solution. We will be covering the topic in detail in the coming articles. Let us start with the basic terminologies:
Machine learning, in layman language, means, to train a machine so that it can learn to think itself. Technically, it is a part of Artificial Intelligence. Giving machines the algorithm to improve their performance automatically with experience. The model is built using the learning algorithm(ML algorithm) and then the model is trained with the help of a dataset. The game is then decided upon the dataset. The better and bigger the dataset, the better the machine’s decision-making capability. Here comes the role of Big Data and why are industries willing to adapt it.
Reinforcement learning is the latest frontier of machine learning.
Deep learning is part of a broader family of machine learning methods based on artificial neural networks with representation learning. It is a machine learning technique that uses artificial neural networks(ANN), which are inspired by biological neural networks. Neural networks and deep learning together have made image recognition, speech recognition, natural language processing(NLP).
Immense positives might come along with negatives such as military use. Collective choices that will be made about AI in the next few years will be for the benefit of all, everyone should take an active role in defining how AI will shape our future.
AI will bring tremendous change in the future, more than these small applications.
There are a number of applications using AI in our daily lives that we don’t even realize. From our mobile phone’s face lock to our social media and Siri and Alexa with speech recognition assistants and much bigger applications including Tesla’s self-driving car. Forbes powerful-examples-of-artificial-intelligence.
Applications in business
Email filtering- sorting out the spam emails
Automatic Language translations
Traffic Prediction- image processing
Stock Market Trading
Can machines think? Alan Turing was the first one to arrive at this question. He is considered the founding father of AI with his discoveries on this subject in 1950. Thereafter, John McCarthy in a conference with summer research interns in Dartmouth came up with the term Artificial Intelligence, 1956. The invention of the first AI language “Lisp” inspired many and it became a language of choice for AI applications.
Basically, Artificial intelligence is a set of algorithms and intelligence to try to mimic human intelligence. The backbone of AI is machine learning.
Applications in business
Human Resource Management
IBM - Deep Blue- On May 11, 1997, an IBM computer called IBM Deep Blue beat the world chess champion after a six-game match. The Deep Blue had the capability to think of 200milion positions per second in Chess. This was the major breakthrough in Artificial Intelligence. Deep Blue had an impact on computing in many different industries. It was programmed to solve the complex, strategic game of chess, so it enabled researchers to explore and understand the limits of massively parallel processing.
Alpha Go - AlphaGo is the first computer program to defeat a Go world champion, and is arguably the strongest Go player in history. Go is a popular game that originated in China involving two players and thinking in multiple layers. It was made by Google, in 2016, using Reinforcement learning and neural networks, which resembles our own decision-making process.
Amazon SageMaker Data Wrangler – Aggregate and Prepare Data for Machine Learning – AWS: This is a business application made by Amazon in 2017. It is a cloud machine learning platform. SageMaker Data Wrangler reduces the time it takes to aggregate and prepares data for machine learning (ML) from weeks to minutes. With SageMaker Data Wrangler, you can simplify the process of data preparation and feature engineering, and complete each step of the data preparation workflow, including data selection, cleaning, exploration, and visualization from a single visual interface. It has continuously updated features since then.
DeepMind: AlphaFold- Human body proteins were a topic of research since the late 1950s. The folded structure of proteins was of great importance because it will give a more thorough understanding of the nature of the disease by investigating the molecules that make up the human body. Google’s deep learning program for the 3D shapes of proteins stands to transform biology. It can now make accurate predictions of what shape a protein will form based on its sequence of amino acids. It won 13th (CASP) Critical Assessment of Techniques for Protein Structure Prediction for its prediction in 2018.
GPT-3: OpenAI’s GPT-3 API allows us to give instructions in plain English and in a result we get the desired task completed. For example, if I write instructions to generate a website design similar to Instagram, it will not only provide the UI but also provide the code for that UI. A college student used GPT-3 to write a fake blog and ended up at the top of Hacker News.
The scope of these technologies is relevant to any intellectual task. Modern artificial intelligence techniques are pervasive and we will be discussing the topic in detail with more focus on an industry sector.
Read more on The Top 10 Technologies of this decade here.
Shriya Madan, an ardent learner, a digital enthusiast, aspires to make an impact in the real world with the power of technology
Pranjali Apurva, driven by curiosity, converging design principles with digital transformation