by Anna Mosolova

Coronavirus brought a lot of changes in our everyday lives affecting all its sectors. New restrictions, protective measures, widespread lockdowns, and mass quarantines had an enormous impact on the industry. In a time when people’s work should be limited and the remaining employees’ activity is to be optimized to the maximum extent, AI is there for you to help with various natural language processing (NLP) tasks. …


by Viktoria Kondrashuk

Keywords:

COVID-19, MANUFACTURING, PRODUCTION PLANNING, COST OPTIMIZATION, LABOR RESOURCES MANAGEMENT, ENTERPRISE RESOURCE PLANNING, ERP, CONVEYOR LINE, AUTOMATION.

Description of the Business Problem. Why Does It Matter?

The pandemic of the COVID-19 virus has affected many sectors of the economy. According to research by the Institute for Supply Management, the manufacturing sector contracted in April 2020, as the Purchasing Managers Index (PMI) registered 41.5 percent, 7.6 percentage points lower than the March 2020 reading of 49.1 percent. The PMI recorded its lowest level since April 2009, when it registered 39.9 percent. The 7.6-percentage point decrease in the PMI is the largest one-month decline since a 9-percentage point decrease in October 2008. Among the big six industries, only food, beverage, and tobacco products expanded. For the second month in a row, all of the PMI subindexes show a strong negative impact due to the ongoing coronavirus pandemic. …


by Viktoria Kondrashuk

Keywords:

COVID-19, PRODUCTION, MATHEMATICAL OPTIMIZATION, SCHEDULE PLAN, SCHEDULE THEORY, SCHEDULE COMPOSITION, COST REDUCTION, LABOR ORGANIZATION, MANAGEMENT OF A PRODUCTION ENTERPRISE, CONVEYOR LINE, AUTOMATION.

Image for post
Image for post

Description of the Business Problem. Why Does it Matter?

This year, the entire world business has faced another crisis; this time was caused by the outbreak of the global pandemic of the new virus COVID-19. The consequences of the virus will affect businesses for many more years and are comparable to the Great Depression or the global financial crisis of 2007–2009.

We discussed working under COVID-19 conditions with business executives to find out what problems production companies faced and how quickly they were able to adapt to new working conditions. …


by Anna Mosolova

The crisis provoked by the coronavirus affected the majority of sectors of the economy. However, some companies not only didn’t experience such a problem, but they also gained profits from this situation because of the enormous increase in demand. One of these industries is educational technologies.

Image for post
Image for post
Relative winning and losing US industries, 2 January to 6 March 2020

Self-isolation and quarantine made people change their educational routines. Primarily students but other persons interested in studying were required to stop attending classes and start taking their courses online. This jump significantly increased demand in the sector of online education. Simplilearn now is serving 3 million users all over the world and is planning to increase this number by 20% during 2020. Vedantu quadrupled its earnings and added 40% more employees this year. …


Image for post
Image for post

by Olga Proka

About preprocessing on the fly challenges

Image data in the real world is usually not in a form that can be directly fed into a neural network.

The amount of high-quality data is often limited. A popular solution to this problem is augmentation. Usually, we should preprocess the data: change the format, remove any defects, align, resize and do other preprocessing actions. Arguably, the most popular tools, in this case, are OpenCV and Pillow because they are very flexible. For example, OpenCV contains more than 2,500 optimized algorithms, including a comprehensive set of both classic and state-of-the-art computer vision and machine learning algorithms.

However, sometimes Data Monsters has a challenge that demands an image to be transformed “on the fly”. Usually, it works with big data in real-time, when you should use really high-speed tools. For example, it could be defect detection at conveyor or the deep learning challenge with the necessity of generating unique samples in real-time. In such cases, the NVIDIA DALI library could be a real salvation. It has three operation modes: in CPU, in GPU and mixed type (accepting data from CPU and producing the output at the GPU side). Recent advances in GPU architectures introduced in the NVIDIA Volta and Turing architectures have significantly increased GPU throughput in deep learning tasks. …


Image for post
Image for post

COVID19 is a big world problem today and in this difficult time and we must be calm and turn to science. Data Monsters are staying home and learning predictive analytics tools about the spread of the virus. High-quality predictions are very important and helpful for the healthcare system because it helps to resolve how many resources must be dedicated to different regions, states, and countries.

So, how we can predict the count of the infected people tomorrow?

The main parameter Δ_(d+1) is the increase in new cases of infection on the new (d+1) day. …


Image for post
Image for post

When you work with computer vision challenges, you must choose a method for measuring the similarity between two images to compare different results of your experiments. Let’s consider different methods in this note.

The most traditional estimator is mean-square error (MSE). MSE measures the average squared difference between the estimated values (predicted values) and the actual value (ground truth). So we just calculate squared differences pixel by pixel. But this works well only if we want to generate a picture with the best pixel colors conformity with the ground truth picture. …


Image for post
Image for post

In 2017, we published our article “25 Chatbot Platforms: A Comparative Table”. Recently we updated the information on chatbot platforms and added 25 more chatbot building tools. We created a new version of the comparative table where we paid our attention to artificial intelligence features, especially to Natural Language Processing, to each platform’s use cases, and to industries in which the tool can be used.

A fragment of the comparative table:


Image for post
Image for post

by Olga Davydova, Data Monsters

In this paper, we will talk about the basic steps of text preprocessing. These steps are needed for transferring text from human language to machine-readable format for further processing. We will also discuss text preprocessing tools.

After a text is obtained, we start with text normalization. Text normalization includes:

  • converting all letters to lower or upper case
  • converting numbers into words or removing numbers
  • removing punctuations, accent marks and other diacritics
  • removing white spaces
  • expanding abbreviations
  • removing stop words, sparse terms, and particular words
  • text canonicalization

We will describe text normalization steps in detail below.

Convert text to lowercase

Example 1. …

About

Data Monsters

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store