Rocky History of Artificial Intelligence

Artificial intelligence is the imitation of action that a human brain performs. This process involves the extraction of information, selecting criteria for using information and seeking for a impactful conclusion.[1] although Computers are created in 1936 and become the integral part of social and professional fabrics in mid 1980’s. AI and big data (the database of the all information) are accepted worldwide in various field of science and technology.

The renowned name Gary kasparov a greatest chess player of all time was conquered by “Deep blue” an IBM trained chess computer.

The classical game show jeopardy was won by IBM trained Watson System by defeating the champions. After this concept was applied to some game shows it was then brought in the field of science and technology. This new from is artificial intelligence whose learning simulate human learning, which can deal human like situation with proper training and can response to any change in data.[2].

Artificial Intelligence has a rocky history spanning back to the 1950s.

Get quality help now

Doctor Jennifer

Verified writer

Proficient in: Artificial Intelligence

5 (893)

“ Thank you so much for accepting my assignment the night before it was due. I look forward to working with you moving forward ”

+84 relevant experts are online

Hire writer

Alan turing was the man behind the idea and crystallization of AI in his paper “Computing Machinery and Intelligence”.The question posed was Can machine think?,In 1956, John McCarthy a computer scientist organised the Dartmouth Conference, at which the term ‘Artificial Intelligence’ was first adopted. Various Research centres across the globe popped up to explore the potential of AI. Researchers Allen Newell and Herbert Simon were the two names that promotes AI as a tool of computer science that could completely change the world’s prospective about use in science and technology.[3].

In the field of medicine AI has subfield i.e machine learning and deep learning. Machine learning(ML)which is having a ability to learn by using Statistical method with or without definite programming. Supervised, unsupervised and reinforcement learning are the characterization of ML [4]–[6]. Deep learning(DL) on the other hand is the type of ML that uses artificial neural network(ANN). DL has achieved a great success in a number of field such as self driven cars, speech recognition, computer vision and drug development.[7] which is clearly described in figure 1. Discovering new generation of information, obtaining higher degree of precision, automated simulation and prediction, diagnosis and detection of disorder and in clinical trails designing are some of the advantages of AI.[8][9]

AI tools helps in predicting in vivo data, pharmacokinetics relationship of a new therapeutics including their quantitative structure-property relationship (QSPR) or quantitative structure-activity relationship (QSAR), fixing the dosing, and skin or blood-brain barrier permeability.[10][11][12] the prediction of pharmacokinetic profile of the pharmacotherapeutic agent, application of in silico model like AI and ANN result in increased output and reduction in the cost of research project.[13] a promising approach i.e neural network modelling which for framing the molecular structure of organic compounds and predicting their physicochemical characteristics.[14] Number of nodes in the input and output layers is determined using the number of independent and dependent variables, respectively. Number of hidden layers and nodes in each layer depend on the complexity of problems. Many ANN models, consist of only one hidden layer, mean- while, more than one hidden layer can be implemented for modeling the complex problems.[15]The sub type of AI, where computers programs (algorithms) learn associations of predictive power from examples in data. ML on the simple language is the implementation of statistical models to data using computers. ML uses a enclose set of statistical techniques than those routinely used in medicine. Novel approaches such as Deep Learning are focused on models with fewer assumptions regarding the underlying data and are therefore are capable to manage a higher number of tedious data.[16]

Machine learning can immensely affect health care and market challenges that produce $100 billion of revenue on annual scale.[17] expert in the field of industry has predicted that drug developed using AI are 2-3 year from launch but in upcoming future but facing critical problem in competing.[18]ML is been implemented in a wide area of health and diagnosis across the globe, other application are in making personalized medicine, drug discovery and manufacturing and clinical trails, radiotherapy and radiology, smart health management bioinformatics, chemical informatics, social network analysis, stock market analysis, and robotics[19] As an example, let’s examine the situation of marking a examinee nucleotide sequence as miRNA or not. One of the most easy approach shall be to determine a set of short nucleotide sequences that are parts of the known miRNA and non-miRNA sequences and to construct a set of rules based on the existence of these nucleotide “words.” For example, one such rule can state that a sequence containing “AGCACU” is more likely to be a miRNA than not. Then one could simply label candidate sequences using these rules. In practice, constructing such a rule based system is very difficult as there are many possible nucleotide words and the mapping is very complex. Instead of manually specifying a complex set of rules, machine learning methods can automatically build a statistical model using these nucleotide words. These models can then be trained using large samples of biological data since the training process is automated. For machine learning, such rules (here a nucleotide hexamer) are determined from features which need to be defined for the input data.