Machine Learning – ISS Art Blog | AI | Machine Learning

Graph Neural Networks and their applications

Jul 18 2022

A Graph Neural Networks (GNN) is a class of artificial neural networks for processing graph data. Here we need to define what a graph is, and a definition is a quite simple – a graph is a set of vertices (nodes) and a set of edges representing the connections between the vertices. There are many data sources which produce data that can be organically presented in a graph form. For example, we can consider social network users as graph vertices where two vertices are connected if corresponding users are friends.

Crowdsourcing in machine learning: expectations and reality

Apr 11 2022

Every person who works in machine learning (ML) sooner or later faces the problem of crowdsourcing. In this article we will try to give answers to the questions: 1) What is in common between crowdsourcing and ML? 2) Is crowdsourcing really necessary?

How much is RMSE? Technical and business metrics – exchange rate.

Mar 18 2022

It is not a secret that people who work in the sphere of Machine Learning sooner or later start to see themselves as a kind of carriers of sacred knowledge. They know what back propagation, cross-entropy, tensors are and how you can transform them.

It goes without saying that staying in this kind of ML bubble affects the perception. The developers start to think that everybody knows about technical specifications of this work, as they do. But it is far from being the truth. In the Machine Learning sphere there are two main parties: engineers and customers. In this article we offer you to see the process of ML development from the point of view of the customer.

Feb 26 2021

Artificial intelligence (AI) is directly related to Data Science – the science about data that aims to extract value from a big mass of information. This value may reside in, for example, enhanced forecasting capabilities, discovering patterns, better decision-making. Basically, AI concerns with information processing algorithms and methodologies. Artificial intelligence operates on huge amounts of data, analyzes it, and uses gathered insight to develop solutions.

Feb 05 2021

We know that efficient supply chains can be a crucial part of differentiating an organization from its competitors. But how can a company truly leverage the advances in artificial intelligence and neural networks we’ve seen in just the last decade?

Particularly for Fast Moving Consumer Goods (FMCG), supply chain trends are emerging that must not be overlooked if a company wishes to become or remain a market leader. Supply chain agility, optimization, sustainability, and ethical considerations challenge established players and give disruptive newcomers opportunities to capture significant market share. This article covers the emerging role of AI in supply chain management. We discuss how it could help companies to more effectively analyze and improve the efficiency of their operations. There is an evolving role advanced technologies play in the FMCG industry. They have potential to dramatically impact supply chain optimization and supply chain productivity.

Nov 11 2020

These days, everyone has heard of artificial intelligence (AI). But not everyone understands what goes into making an AI algorithm work properly. At a high level, AI (also known as machine learning), works by ingesting a large set of data called training data. Then, the AI uses an algorithm to sort through this data and discover trends. This algorithm can be a bit of a black box, and there are many variations of machine learning algorithms. But the important part to remember is that AI “trains” on a dataset. This dataset is called training data. After an AI is fully trained on a dataset, it can be applied to “test” or “application” data. That is where you will typically see AI in action.

Of those three phases (the training data, algorithm, and test data), today we will focus on training data. Specifically, we will discuss how training data is collected today, introduce an alternative called synthetic data, and examine the several advantages of synthetic data.

Apr 27 2020

We hear a lot of buzz around some of the more popular emerging technologies like artificial intelligence and machine learning and for good reason. But, there are a few others that work in conjunction with these technologies that are on the verge of completely changing how we look at both macro and microeconomics.

Examining each one of these technologies individually is inspiring on their own. Perhaps even more awe-inspiring is when we consider in what ways they can work together and what that may mean for the world. One thing is for certain—the name of the game in the 2020s is disruption and the ways we will find to combine technologies this decade mean the world as we know it is about to fade into history.

Apr 09 2020

A recent study has shown that a lack of skilled talent is the top hiring challenge, as 87% of HR professionals say that there were few qualified applicants for the positions they were trying to fill.

This is the reason why there’s currently a war for talent raging on, and companies have to go out of their way in order to attract and keep skilled people.

The screening and hiring processes are also complex and challenging. Companies receive avalanches of CVs for every vacant position, and it’s time-consuming to review every single one of them.

However, thanks to artificial intelligence (AI) and HR software, it’s much easier to find and hire the right person for the job.

Mar 04 2020

Introduction

Cracks on the surface are a major defect in concrete structures. Early crack detection allows preventing possible damage. There are various approaches to solving this problem. It can be manual inspection or automatic detection methods. But nowadays automatic detection methods include not only laser testing and radiographic testing. Progress in neural networks and computer vision allows us to use image processing for concrete surface crack detection.

In this article, we will share our approach to solving the problem mentioned above.

Dec 25 2019

Implementing DialogFlow chatbots is cool and convenient if you have something trivial and easy to prototype: fancy UI – easily, extracting base entities like name, surname and phone number – here is a tool if don’t want to install and deploy – cloud solution is at your service.
But what if you need to go deeper:

– Do you have a Japanese tokenizer, dear DialogFlow?
– Nope
– Transparent and customizable intent classification tool?
– Sorry, guys.
– Also I want to integrate my search index, knowledge graph and custom dialogue management policy.
– What are you talking about?

In a nutshell if you want fully controlled system, if you need custom advanced AI in your app, if you need natural language processing in your chatbot pipelines, if you want to scale your chatbot behaviour – on-premise solution is the way for a chatbot developer. And here it’s Rasa framework that really shines.

Category - Machine Learning

Graph Neural Networks and their applications

Crowdsourcing in machine learning: expectations and reality

How much is RMSE? Technical and business metrics – exchange rate.

5 AI Revolution Benefits Entrepreneurs Should Look Forward To

AI Predictive Supply Chain Management

What are the advantages of synthetic data?

4 Ways Machine Learning Is Disrupting the FinTech World in 2020

5 Ways to Improve Your Recruiting Strategy With AI and HR Software

Concrete crack detection using Deep Learning and Computer Vision

Introduction

Migration From DialogFlow to RASA: the missing part

Contact us

Categories