GROUP 2 - NEURAL NETWORK (2nd reporter).pdf

Neural Networks next slide 02 thynk unlimited Let’s walk through What exactly is a Neural Network is. The major types and how they compare with each other, and A brief summary of other Neural Networks that you...

Neural Networks next slide 02 thynk unlimited Let’s walk through What exactly is a Neural Network is. The major types and how they compare with each other, and A brief summary of other Neural Networks that you can use. next slide  thynk unlimited 03 Neural Networks A Neural Network is a working system at the heart of a Deep Learning algorithm that helps it process raw data. In fact, Neural Networks are pretty much like the human brain. Like our brain that’s formed of a network of neurons, Neural Networks are based on connected nodes or units – technically known as artificial neurons. Their job? To identify hidden correlations and patterns in raw data, classify them, and continuously improve. Neutral networks sit at the center of all Deep Learning algorithms. So to understand a model or create one, you need to first dive into the architecture of a Neural Network. next slide  thynk unlimited 04 Types of Neural Networks ANN RNN Artificial Neural Networks Recurrent Neural Networks synonymous with Neural are unique on account of their Networks. However, both ability to process both past data aren’t one and the same. and input data — and memorize Instead, an ANN is a type things — and were developed to of Neural Network. overcome the weaknesses of the feed-forward network. Practical applications include Google’s voice search and Apple's Siri. next slide Artificial Neural Networks (ANN)  ANN Artificial Neural Networks contain artificial neurons which are called units. These units are arranged in a series of layers that together constitute the whole Artificial Neural Network in a system. A layer can have only a dozen units or millions of units as this depends on how the complex neural networks will be required to learn the hidden patterns in the datase  How it works? Commonly, Artificial Neural Network has an input layer, an output layer as well as hidden layers. The input layer receives data from the outside world which the neural network needs to analyze or learn about. Then this data passes through one or multiple hidden layers that transform the input into data that is valuable for the output layer. Finally, the output layer provides an output in the form of a response of the Artificial Neural Networks to input data provided. ANN ANN Architecture In the majority of neural networks, units are interconnected from one layer to another. Each of these connections has weights that determine the influence of one unit on another unit. As the data transfers from one unit to another, the neural network learns more and more about the data which eventually results in an output from the output layer ANN Artificial neurons vs Biological neurons  ANN TYPES Feedforward Neural Network: A basic neural network where data flows in one direction, from input to output. Hidden layers may exist, but there’s no backpropagation. Convolutional Neural Network (CNN): Similar to a feedforward network but includes convolutional layers for processing input. Used widely in speech and image processing, particularly in computer vision. Modular Neural Network: A system of independent neural networks that handle different tasks with no interaction. It simplifies complex problems by breaking them into smaller tasks.  ANN TYPES Radial Basis Function Neural Network: Comprises two layers, where the input is mapped to radial basis functions and then passed to the output layer. Commonly used for modeling data trends. Recurrent Neural Network (RNN): Uses outputs from previous layers as inputs for better prediction. Each unit retains memory of previous steps, aiding in sequential data processing.  Real Life Applications Healthcare: In oncology, ANNs help detect cancerous tissue with the same accuracy as doctors. Facial analysis using ANNs can also spot rare diseases early, improving diagnostics and healthcare quality. Personal Assistants: Siri, Alexa, and Cortana use speech recognition powered by ANNs and Natural Language Processing (NLP) to understand and respond to user queries, managing language syntax and conversation context.  Real Life Applications Social Media: Artificial Neural Networks (ANNs) power features like Facebook's "People you may know" by analyzing profiles, interests, and connections. ANNs also drive facial recognition, identifying key reference points on faces and matching them in the database using convolutional neural networks (CNNs). Marketing and Sales: E-commerce platforms like Amazon and Zomato use ANNs for personalized recommendations based on browsing history and preferences, tailoring marketing campaigns across various industries like books, movies, and hospitality. 0X ADVANTAGES & DISADVANTAGES of ANN next slide SAGUN Advantages Ability to handle complex data Non-linear modeling capabilities Adaptability and learning capabilities Robustness to noisy or incomplete data Feature extraction capabilities Domain agnostic, applicable to various business areas Parallel processing for efficient computation Can handle high-dimensional data Can uncover hidden patterns and insights next slide SAGUN Disadvantages Need for large amounts of labeled training data Black box nature can hinder interpretability Computationally intensive and resource-consuming Lack of transparency in decision-making Potential overfitting without proper regularization Complexity and difficulty in model tuning Difficulty in explaining results to stakeholders Sensitivity to input data quality and preprocessing Ethical considerations in sensitive decision-making next slide SAGUN Applicaiton of ANN(Medical Field) SAGUN Recurrent Neural Networks (RNN) Overview 08 Introduction to RNN How RNNs Work Applications of RNNs Advantages of RNNs Challenges of RNNs Conclusion next slide RECURRENT NEURAL NETWORK (RNN) Recurrent Neural Network(RNN) is a type of neural network designed for processing sequences of data. Unlike traditional neural networks, RNNs have connections that form directed cycles, allowing them to maintain a "memory" of previous inputs through their hidden states. This feature enables RNNs to recognize patterns in data where context and order matter, making them ideal for tasks involving sequential data Recurrent Unit and Memory Update The recurrent unit is the basic building block of an RNN. Think of it as a small decision-making unit that processes information step by step. At each time step, this unit takes two inputs: 1. Current Input: The data or information you want the RNN to process at that moment (e.g., a word in a sentence). 2. Previous Memory (Hidden State): This is the "memory" of what the RNN has already seen or learned from previous inputs. Types Of RNN Variations Of Recurrent Neural Network (RNN) Bidirectional Neural Network (BiNN) - A type of neural network that processes data in two directions: from the past to the future and from the future to the past. Long Short-Term Memory (LSTM) - A type of neural network designed to remember information for a long time and forget irrelevant information. It’s a specific kind of Recurrent Neural Network (RNN) that handles sequences of data. Applications of Recurrent Neural Network 1. Language Modelling and Generating Text 2. Speech Recognition 3. Machine Translation 4. Image Recognition, Face detection 5. Time series Forecasting  Advantages of RNN (Recurrent Neural Networks) 12 Handles Sequence Data Well: RNNs are great for tasks where order matters. Remembers Past Information: RNNs can remember past data to improve predictions. Efficient: RNNs don’t need new rules for every word or data point, making them more efficient. Works with Any Length of Data: Whether analyzing a short text message or a long email, RNNs can handle both lengths.  Challenges of RNN Forgets Over Long Sequences: Think of your phone’s autocorrect. If you’ve been typing a long message, it might not always remember what you wrote earlier and suggest something that doesn’t make sense based on the start of your message. Slow to Learn: RNNs take time to learn from data. Limited Memory: RNNs may have trouble recalling earlier parts of a long conversation. Hard to Adjust: Adjusting an RNN to work perfectly can be as tricky as fine- tuning a complicated machine.  thynk unlimited Thank You! Thank you for exploring the world of AI with us. For more information, please contact us at [email protected] +123-456-7890 www.reallygreatsite.com