AI This document consists of 19 printed pages. Subject Code: 9618 Exam Date: 25/5/2024 Topics Included CH18 - Artificial Intelligence (ai) Instructions • noooo. Student Details (must be filled) Name Jenish Jung Thapa ________________ Roll No _______ Marking (for teachers only) This worksheet includes 15 questions. Total ____/____ Grade ____ Signature Date Completed __________ ___/___/______ Grade/Div ______/_____ The learning process that uses artificial neural networks that contains high 9618_w22_qp_33/Q9 1 number of hidden layers thats modelled according to the human brain. Deep learning uses many layers to progressively extract higher level features from the input. Its specialized form of machine learning. It makes good use of unstructured data. Deep learning outperforms other methods if data size is large.It enables machine to process data with a nonlinear approach. It is effective at finding hidden patterns,patterns that human can't find out, too complex patterns.It can provide more accurate outcome with more no of hidden layers. Supervised learning allows data to be collected, or to produce output based on previous experiences. It uses labelled input data.In supervised learning, known input and associated outputs are given to train the machine. So, it is able to predict future outcomes based on past data. Unsupervised machine learning helps all kind of unknown patterns in data to be found.It only requires input data to be given. It uses unlabelled input data. Its not trained on right output. Generated using Markhint (markhint.in) 9618_w22_qp_32/Q7 2 9618_w21_qp_31/Q9 5 It enables deep learning to take place.When the problem you require to solve has higher complexity it requires more layers to solve.To enable the neural network to learn and make decisions on its own.To improve the accuracy of the result. Artificial neural networks are intended to replicate the way human brain works. The values are assigned for each connection between nodes. The data are input into input layer and are passed into system. The data are analyzed at each subsequent layer where characteristics are extracted and outputs are calculated. This process of learning and training is repeated many times to achieve optimum output i.e. reinforcement learning takes place. Decisions can be made without being specifically programmed. The deep learning network will have created complex feature detectors. The output layer provide the results. Back propagation of errors will be used to correct any errors that have been made. Generated using Markhint (markhint.in) 9618_w21_qp_31/Q9 5 B C 12 7 4 9 4 7 A-E 7 3 10 A-E-F 8 3 11 A-B 12 A-E-SCHOOL A-E-F-SCHOOL 11 0 11 HOME-A-E-F-SCHOOL. Generated using Markhint (markhint.in) 13 0 12 11 9618_s21_qp_33/Q5 6 3 5 Generated using Markhint (markhint.in) 2 9 3 8 between nodes. AI can be defined as finding path in a graph. Graph can be analyzed using 9618_s21_qp_31/Q5 Artificial neural network can be represented using graph. The graph can provide the relationship different algorithms as dijkstra or a* algorithm. Then back propagation and regression can be used to predict the future outcomes by just looking at the input. 9618_s23_qp_33/Q2 9 The main purpose of Dijkstras and A* algorithm is to find the optimal,shortest and most cost effective route between two nodes based on the distance,cost and time. Deep learning,Supervised,unsupervised,reinforcement Generated using Markhint (markhint.in) 9618_s23_qp_32/Q10 10 9618_s23_qp_32/Q10 begin begin begin C B 6 C 5 13 7 12 15 D 4 11 G 7 5 F G F 7 END 11 12 BEGIN-C-G-F-END. Generated using Markhint (markhint.in) 1 0 12 12 12