Artificial Intelligence

AI / ML Overview

Artificial Intelligence vs Machine Learning vs Deep Learning

Although the three terminologies are usually used interchangeably, they do not quite refer to the same things.

ArtificiaI Intelligence is the broader field focused on building applications with the ability to make decisions and perform tasks that simulate human intelligence and behavior. What are the Real-world benefits of AI? In health care, treatment effectiveness can be more quickly determined. In retail, add-on items can be more quickly suggested. In finance, fraud can be prevented instead of just detected. And so much more. (Drive Revenue, Reduce Costs, Avoid Risks)

Machine learning is, more or less, a way for computers to learn things without being specifically programmed. But how does that actually happen? The answer is, in one word, algorithms. Algorithms are sets of rules that a computer is able to follow.

Deep learning is a type of machine learning that can process a wider range of data resources, requires less data preprocessing by humans, and can often produce more accurate results than traditional machine-learning approaches. In deep learning, interconnected layers of software-based calculators known as “neurons” form a neural network. The network can ingest vast amounts of input data and process them through multiple layers that learn increasingly complex features of the data at each layer. The network can then make a determination about the data, learn if its determination is correct, and use what it has learned to make determinations about new data. For example, once it learns what an object looks like, it can recognize the object in a new image.

Generative AI refers to a type of artificial intelligence system that is designed to generate new, original content, such as images, text, music, or even entire videos. Unlike other AI systems that are primarily focused on analyzing and interpreting existing data, generative AI models are capable of creating new data based on patterns and examples they have learned during the training process. Generative AI models are typically built using deep learning techniques.

Traditional Programming vs Machine Learning

In traditional programming, code is developed to process data, apply rules, to identify answers or the desired output. In Machine Learning, rather than receiving explicit programming instructions, the Machine learning algorithms detect patterns and learn how to make predictions and recommendations by processing data and experiences. Through an iterative process, Algorithms are tuned (parameters optimized) to achieve a resulting model that can adapt in response to new data and experiences to improve efficacy over time.

Model Development Life-Cycle

Business Understanding: What is the business need and expected value to be achieved? Understanding the business goals helps determine what we want to do and how to know if we are successful.

Data Understanding & Preparation: Understanding the data (EDA: Exploratory Data Analysis) and preparing/converting it into forms that will yield the best results. Create a training and validation training set. Imputation fills in the missing values with some number. For instance, we can fill in the mean value along each column. Pipelines are a simple way to keep your data preprocessing and modeling code organized. Specifically, a pipeline bundles preprocessing and modeling steps so you can use the whole bundle as if it were a single step. Data leakage (or leakage) happens when your training data contains information about the target, but similar data will not be available when the model is used for prediction.

Modeling: Data mining techniques are applied to the data in an iterative fashion to identify the solution that best achieves the business goals. Define a model (select algorithm, fit a model (train) looking for patterns in the data, make predictions, validate the model.

Evaluation: Once the model is optimized, the results are reviewed to determine if they are valid and reliable for deployment.

AI Solution Development: Deploying a ML model into production can take on many forms, from a simple API to a complete supporting application.

Featured video

Page updated

Report abuse