Day 1 #SheBuildsOnAzure

Microsoft Azure#Include<Her> Cohort 2.0

Key Highlights

• Fundamentals of AI and ML
• Classification of ML (Supervised Learning, Unsupervised Learning and Reinforcement Learning)
• ’UCF‘ — User Collaborative Filtering

Artificial intelligence is the simulation of human intelligence processes by machines, especially computer systems. Specific applications of AI include expert systems, natural language processing, speech recognition and machine vision. The Azure AI-900 session had equipped us with the fundamental knowledge and skills required to understand and leverage AI capabilities on the Azure platform. It has served as a stepping stone for further exploration of advanced AI concepts and Azure AI services.

Benefits of Using Cloud Scalability: Cloud services provide the ability to scale resources up or down according to demand. This flexibility allows businesses to easily accommodate fluctuations in traffic, storage needs, and computational requirements without having to invest in and manage physical infrastructure.

Cost Efficiency: Cloud computing eliminates the need for upfront capital expenditures on hardware and infrastructure. Accessibility and Availability: Cloud services are accessible over the internet, enabling users to access their data and applications from anywhere, at any time, and from any device with an internet connection. We got to know that Microsoft currently has three data centers in India, located in Mumbai, Chennai, and Pune. Additionally, we learned that a fourth data centre is proposed for Hyderabad, with the growing presence of cloud infrastructure in the country. Creating Resource Groups We got to know about how the cloud resources works practically. We learned how to create a resource group in Azure after obtaining a subscription. This step is important for organizing and managing your cloud resources efficiently. AI Services in Microsoft Azure Then we got an insight about the Artificial Intelligence (AI) services offered by Microsoft Azure. We explored various AI services, including Azure ML (Machine Learning), Cognitive Services, Azure Bot Services, and Azure Cognitive Search. These services empower developers to leverage AI capabilities and enhance their applications with intelligent features. Dataset In the field of AI and ML, a dataset refers to a collection of data that is used to train, validate, or test machine learning models. A dataset contains samples or examples, each representing a data point with various features or attributes. For example, in a dataset of images, each image is a data point, and the features could include pixel values, labels, or metadata. Datasets are crucial for training AI models as they provide the necessary information for the models to learn patterns, make predictions, or perform tasks. High-quality and diverse datasets are essential for achieving accurate and reliable AI models.

Machine Learning can be divided into three categories which include: ◾Supervised Machine Learning

◾Unsupervised Machine Learning

◾Reinforcement Machine Learning + These three categories represent different approaches to machine learning, each suited for specific types of problems and data availability. It’s worth noting that there are also hybrid approaches that combine elements from multiple categories, such as semi-supervised learning (a combination of supervised and unsupervised learning) and transfer learning (leveraging knowledge from one task to improve performance on another).

+ Supervised Learning: Supervised learning involves training machine learning models on labeled datasets. In this approach, the dataset consists of input data along with corresponding output labels or target values.

+ Unsupervised Learning: Unsupervised learning deals with datasets that do not have predefined labels or target values. The goal is to discover patterns, structures, or relationships in the data without explicit guidance. + Reinforcement Learning: Reinforcement learning involves training agents to interact with an environment and learn optimal actions based on feedback and rewards. The agent learns through trial and error, receiving positive or negative rewards for its actions. ⃣

SUPERVISED MACHINE LEARNING

▪️ Has quite good knowledge of the data.

▪️ Works on well-labeled datasets. ⃣

UNSUPERVISED MACHINE LEARNING

▪️ Have no prior information regarding the dataset.

REINFORCEMENT LEARNING

▪️ Its performance gets better over time.

▪️ Have no prior information regarding the dataset. Stock Market Analysis is a time series analysis. (This is a supervised Machine Learning). Difference between Unsupervised and Reinforcement ML

▪️ Performance of Unsupervised ML doesn’t get better over time. ▪️ Reinforcing ML works on a feedback mechanism and also tries to upgrade its performance over time. Ground Truth ( X-Ray has L1 cancer)

◾False positive.

◾True positive.

◾True negative (Have L4 cancer).

◾False negative (Doesn’t have cancer).

If expectations matched or not — True/False. If ground truth matched or not — positive/negative. Also learned about various datasets which include COCO dataset, EGO HANDS dataset and Yahoo Finance. Learned about ‘UCF’ which stands for User Collaborative Filtering, the technique behind the Recommendations algorithm of various platforms such as YouTube, Netflix, etc