ARTIFICIAL INTELLIGENCE PRESENTATION Learning, Genetics, and Neural Networks Team Members Kanishk Bhadauria 01214802719 Muskan Garg 02614802719 Nikhil Jha 03414802719 Aparna Jha 03914802719 Learning Leraning is done by viewing listening, interactions, studying and by experience. Leeearing providesus the power to reason, ability to handle new situations and enables us to act in an intelligent way. Human beings are intelligent as they possess knowledge of world. Similarly, making a machine intelligent means it should have the power of learning. Leraning is essentialfor unknown environments Learning is uselful as a system construction method.. Learning modifies the agent's decision mechanism to improve performance There are three general categories of learning that artificial intelligence & machine learning utilizes to actually learn Supervised Learning: The machine has a “teacher” who guides it by providing sample inputs along with the desired output. The machine then maps the inputs and the outputs. This is similar to how we teach very young children with picture books. Almost all of the AI machines we have today have used this form of learning (from speech recognition to self-driving cars). Unsupervised Learning: This is the most important and most difficult type of learning and would be better titled Predictive Learning. In this case the machine is not given any labels for its inputs and needs to “figure out” the structure on its own. This is similar to how babies learn early in life. For example they learn that if an object in space is not supported it will fall Reinforcement Learning: It is defined as “a computer program interacts with a dynamic environment in which it must perform a certain goal (such as driving a vehicle), without a teacher explicitly telling it whether it has come close to its goal. “. The agent is awarded for good responses and punished for bad ones. Methods of Learning in context of machines are as follow Rote Learning: Rote learning is the process of memorizing specific new items as they are encountered. The basic idea is simple and easy to realize within a computer program. Each time a new and useful piece of information is encountered, it is stored away for future use. Learning by taking advice: Both humans as well as machines can learn through advice. This type is the easiest and simple way of learning. In this type of learning, a programmer writes a program to give some instructions to perform a task to the computer. Once it is learned (i.e. programmed), the system will be able to do new things. Learning by induction: Inductive learning involves using evidence to determine the outcome. Inductive reasoning refers to using specific cases to determine general outcomes, e.g. specific to general. Learning by Deduction (or Relevance-based learning): Unlike inductive learning, which is based on the generalization of specific facts, deductive learning uses the already available facts and information in order to give a valid conclusion. It uses a top-down approach. The one major thing to note is that in deductive learning, the results are certain i.e, it is either yes or no. Whereas it’s probability-based on inductive learning i.e, it can range from strong to weak. Since, deductive reasoning works on preavailable logical facts, let’s have a look. a. All carnivores eat meat. b. Lion is a carnivore. Conclusion: – Lion eats meat. Learning by Analogy: The analogy method in AI is teaching a machine a new topic by connecting it with familiar information. For example, we might rely on portraying the white blood cells of our body as soldiers in our defense system. Explanation Based Learning: Explanation based learning has ability to learn from a single training instance. Instead of taking more examples the explanation based learning is emphasized to learn a single, specific example. Genetic Algorithm Genetic Algorithm is one of the heuristic algorithms. They are used to solve optimization problems. They are inspired by Darwin’s Theory of Evolution. They are an intelligent exploitation of a random search. Although randomized, Genetic Algorithms are by no means random. Genetic Algorithm works in the following steps Step 1 Randomly Step 2 generate a set of Using a fitness function, test possible solutions to a problem. each possible solution Represent each solution as a fixed against length character string. evaluate them. Step 3 Keep the best solutions. Use best solutions to generate new possible solutions. the problem to Step 4 Repeat the previous two steps until either an acceptable solution is found or until the algorithm has completed iterations through a its given number of cycles / generations Basic Operators The basic operators of Genetic Algorithm are 1. Selection (Reproduction) It is the first operator applied on the population. It selects the chromosomes from the population of parents to cross over and produce offspring. It is based on evolution theory of “Survival of the fittest” given by Darwin. There are many techniques for reproduction or selection operator such as: 1. Tournament selection 2. Ranked position selection 3. Steady state selection etc. 2. Cross Cover Population gets enriched with better individuals after reproduction phase. Then crossover operator is applied to the mating pool to create better strings. Crossover operator makes clones of good strings but does not create new ones. By recombining good individuals, the process is likely to create even better individuals 3. Mutation Mutation is a background operator. Mutation of a bit includes flipping it by changing 0 to 1 and vice-versa. After crossover, the mutation operator subjects the strings to mutation. It facilitates a sudden change in a gene within a chromosome. Thus, it allows the algorithm to see for the solution far away from the current ones. It guarantees that the search algorithm is not trapped on a local optimum. Its purpose is to prevent premature convergence and maintain diversity within the population. Advantages Genetic Algorithms offer the following advantages Point 1 Point 2 Genetic Algorithms are better than conventional AI. This is because they are more robust. They do not break easily unlike older AI systems. They do not break easily even in the presence of reasonable Point 3 While performing search in multi modal state-space or large state-space, Genetic algorithms has significant benefits over other typical search optimization techniques. noise or if the inputs get change slightly. Flow Chart The following flowchart represents how a genetic algorithm works Neural Networks Neural networks are a series of the They are used in a variety of operations of an animal brain to applications in financial services, recognize relationships between from forecasting and marketing vast amounts of data. research to fraud detection and risk algorithms that mimic assessment. Neural process networks layers with are several known as As such, they tend to resemble the "deep" networks and are used for connections of neurons deep learning algorithms synapses found in the brain. and A simple Neural Network Input layer Hidden layer Output layer Convolutional Neural Network: A convolutional neural network is one adapted for analyzing and identifying visual data such as digital images or photographs. Recurrent Neural Network: A recurrent neural network is one adapted for analyzing time series data, event history, or temporal ordering. Deep Neural Network: Also known as a deep learning network, a deep neural network, at its most basic, is one that involves two or more processing layers. Applications of Neural Networks Neural Networks are regulating some key sectors including finance, healthcare, and automotive. As these artificial neurons function in a way similar to the human brain. Here are some of the key applications of Neural Networks Facial Recognition. Stock Market Prediction. Social Media. Aerospace. Defence. Healthcare. Signature Verification and Handwriting Analysis.