Artificial Intelligence and it's usage explained

Artificial Intelligence

1. Artificial Intelligence

2. What is AI The study of computer systems that attempt to model and apply the intelligence of the human mind. A branch of computer science dealing with the simulation of intelligent behavior in computers. The capability of a machine to imitate intelligent human behavior. Ability to interact with the world (speech, vision, motion, manipulation) Ability to model the world and to reason about it Ability to learn and to adapt To build systems that exhibit intelligent behavior To understand intelligence in order to model it 2

3. Early History In 1950 English mathematician Alan Turing wrote a landmark paper titled “Computing Machinery and Intelligence” that asked the question: “Can machines think?” 1 2 3 4 5 6 3

4. Computer Chess 2/96: Kasparov vs Deep Blue Kasparov victorious: 3 wins, 2 draws, 1 loss 3/97: Kasparov vs Deeper Blue First match won against world champion 512 processors: 200 million chess positions per second How do you think it works??? 1 2 3 4 5 6 4

5. Applications of AI Various un-realistic applications of AI in real life. Predicting future. (Earth quakes, Tsunamis) Chat Bot (Virtual personal assistants like Siri, Google Assistant etc.,) Self driving cars. (Ex: Tesla - a blessing for elderly and disabled.) AI Eye Doctor (Google’s initiative with Indian eye care chain which examines retinal scans and diagnoses Diabetic retinopathy which causes blindness.) AI Music Composer (Using Deep learning). AI Dream Machine (Dream reading machine in the form of video, Close to 60% accuracy) 1 2 3 4 5 6 5

6. 6 Common Buzz words in AI Computer Science Image Processing Artificial Intelligence Symbolic Learning Statistical Learning Computer Vision Robotics RNN CNN Speech Recognition Computer Vision NLP Machine Learning Pattern Recognition Prediction Classification Unsupervised Learning Reinforcement Learning Supervised Learning Deep Learning(NN)

7. Machine Learning A step towards Machine Artificial Intelligence.. 7

8. Machine Learning Problem Statement: Determine the species of the flower. ML provide computers the ability to learn without explicitly being programmed. Train the machine on training dataset which is large enough to create a model, which helps the machine to take decision based on its learning. 1 2 3 4 5 6 8

9. Machine Learning Decision Making 1 2 3 4 5 6 9

10. Types of Machine Learning 1 2 3 4 5 6 10 Supervised Learning Unsupervised Learning Reinforcement Learning

11. Supervised Learning 1 2 3 4 5 6 11 Here the human experts acts as the teacher where we feed the computer with training data containing the input/predictors and we show it the correct answers (output) and from the data the computer should be able to learn the patterns. Supervised learning algorithms try to  model relationships and dependencies between the target prediction output and the input features  such that we can predict the output values for new data based on those relationships which it learned from the previous data sets. List of Common Algorithms Nearest Neighbor Naive Bayes Decision Trees Linear Regression Support Vector Machines (SVM) Neural Networks

12. Unsupervised Learning 1 2 3 4 5 6 12 The computer is trained with unlabeled data. Here there’s no teacher at all, actually the computer might be able to teach you new things after it learns patterns in data, these algorithms a particularly useful in cases where the human expert doesn’t know what to look for in the data. It is mainly used in pattern detection and descriptive modeling. However, there are no output categories or labels here based on which the algorithm can try to model relationships. These algorithms try to use techniques on the input data to mine for rules, detect patterns, and summarize and group the data points which help in deriving meaningful insights and describe the data better to the users. List of Common Algorithms k-means clustering, Association Rules

13. Reinforcement Learning 1 2 3 4 5 6 13 Reinforcement learning algorithm (called the agent) continuously learns from the environment in an iterative fashion. In the process, the agent learns from its experiences of the environment until it explores the full range of possible states. Reinforcement Learning is a type of Machine Learning, and thereby also a branch of Artificial Intelligence. It allows machines and software agents to automatically determine the ideal behavior within a specific context, in order to maximize its performance. Simple reward feedback is required for the agent to learn its behavior; this is known as the reinforcement signal. List of Common Algorithms Q-Learning Temporal Difference (TD) Deep Adversarial Networks Use cases: Some applications of the reinforcement learning algorithms are computer played board games (Chess, Go), robotic hands, and self-driving cars.

14. Limitations of Machine Learning 1 2 3 4 5 6 14 Are not useful while working with high dimensional data (Large number of inputs and outputs). Cannot solve crucial AI problems like NLP, Image recognition etc., which are having high dimensional data. One of the biggest challenges with traditional ML models is a process called “Feature extraction”. For complex problems like object recognition or handwriting recognition, this is a huge challenge.

15. Deep Learning A step towards Machine Artificial Intelligence.. 15

16. Deep Learning To the rescue... The concept of Deep learning came on late 1950’s, but it was kept on hold due to high-end resource capability. Deep learning is implemented through Deep Neural Network. Motivation behind Neural networks is the biological neuron. Dendrite  Preceptrons (Perceives the input). Axon  Weightages (W1..Wn). 1 2 3 4 5 6 16

17. What is Deep Learning Deep Learning uses Deep Networks which are nothing but Neural Network with multiple hidden layers. The no. of hidden layers can be decided based on the use case and it is possible with the help of advancements in high resource capabilities. 1 2 3 4 5 6 17

18. Deep Learning Input layer – Identifies Dimensionality, Consumes the entire image and identifies the edges of the images through contrast levels. Hidden layer1 – Identifies face features like eyes, nose etc., Hidden layer2 – Able to form the entire face. Output layer – Classifies the images and given a name. Think if any of the layer is missing? 1 2 3 4 5 6 18 Use case1: Image Recognition

19. Deep Learning Multi layer perceptron process the image on pixel by pixel and provides an output. 1 2 3 4 5 6 19 Use case1: Handwritten digits

20. Applications of Deep Learning 1 2 3 4 5 6 20 Facebook’s Face Recognition. Earlier, Face book just identifies the human face and ask us to tag it. But now, it can able identify who it is. Possible with the help of Deep learning and Facebook’s DL library – Caffe2.

21. Deep Learning Point your phone at a flower and it can tell you what type of flower it is. 1 2 3 4 5 6 21 Google Lens

22. How to implement Deep Learning Deep Learning can be implemented using Python language and Tensorflow – a python package to implement Deep learning models. 1 2 3 4 5 6 22

23. Difference between ML and DL 1 2 3 4 5 6 23

24. Emerging Edge AI applications 1 2 3 4 5 6 24

25. Convolutional Neural Networks (CNN) For Analysing images… 25

26. Convolutional Neural Network Input layer – Identifies Dimensionality, Consumes the entire image and identifies the edges of the images through contrast levels. Hidden layer1 – Identifies face features like eyes, nose etc., Hidden layer2 – Able to form the entire face. Output layer – Classifies the images and given a name. Think if any of the layer is missing? 1 2 3 4 5 6 26 Use case1: Image Recognition

27. Recurrent Neural Networks (RNN) For Analysing images… 27

28. Recurrent Neural Network Imagine you have a normal feed-forward neural network and give it the word „neuron“ as an input and it processes the word character by character. At the time it reaches the character „r“, it has already forgotten about „n“, „e“ and „u“, which makes it almost impossible for this type of neural network to predict what character would come next. A Recurrent Neural Network is able to remember exactly that, because of it’s internal memory. It produces output, copies that output and loops it back into the network. 1 2 3 4 5 6 28

29. Applications of RNN Can be used with NLP to predict the next word in a sentence. Speech Recognition. Advancement – LSTM NN. 1 2 3 4 5 6 29

30. Caption lorem ipsum 30 Video slide

31. Caption lorem ipsum 31 Video slide

32. Thank You Manikandan Rangaraj [email protected]