“Since the Vancouver police use predictive modelling to fight crime, it is a great city to gather crime prediction experts and confront theory with practice” says Patryk Miziuła, data scientist and lead instructor at deepsense.ai. Miziula is slated to be a panelist at the Joint Statistical Meetings 2018 conference held in Vancouver on July 28 – August 2.

Crime forecasting is usually associated with the film “Minority Report”, based on Philip K. Dick’s short story. It depicts a dystopian vision of people convicted of a crime before actually committing it.  In the story, information about potential killers or burglars comes from “precogs”, people with precognition talents who saw the future crimes.
“Fortunately, statistical data collected by the police, surveillance systems and municipal services is enough to build a system that can predict the probability of crimes in the area,” says Miziuła. “deepsense.ai prepared such a system for Portland, Oregon and we see a similar solution working efficiently every day in Vancouver”.

Reaching for PreCrime

The Vancouver crime prediction system was introduced in August 2017, following six months of testing in 2016. The pilot ran from April 1 to September 30, when people tend to go on vacation and leave their houses vulnerable to burglaries. During the testing time police recorded a 27% drop in break-ins due to the better distribution of police patrols.
According to Vancouver Police Department data, the total crime rate in the city decreased slightly (-1,5%) in 2016-2017, with the burglary rate falling the most  (18.4%).
The crimes are predicted by a machine learning model, which analyzes the data collected by the city police department and looks for patterns in crimes.
“Machine learning is currently the best tool to handle the massive amounts of data generated on crime. The experiment we did in Portland showed that bad neighborhoods tend to remain bad and crimes usually occur within close proximity. Given that, statistics and machine learning are great tools not only for predicting crime, but also for spotting the problem areas and inspire other public services to act. If vandalism is common in the district, maybe it is not only a matter of crime, but also suggests the need to implement social policy”, Miziuła explained. Further details of the experiment, the results and the technology used in the Portland predictive model are described in a blog post on deepsense.ai’s blog. The experiment will also be described during the upcoming Joint Statistical Meetings 2018 conference held in Vancouver on July 28 – August 2.

Datagazers

The model was built for a US National Institute of Justice competition. The crimes were split into a few categories – burglary, car theft and street crimes including assaults, robberies and shots fired. To predict the place and time a crime would probably occur, the Institute provided historical data with all the crimes registered in Portland between March 2012 and February 2017. deepsense.ai researchers fed the data to its machine learning model, which analyzed the dataset.
“During my speech I will discuss the details of our solution and show how statistics is the best fuel for machine learning techniques,” Miziuła says. “We are essentially datagazers, who can see the future by looking at data, rather than the stars”.
“Vancouver provides a great environment for a conference with crime forecasting on the agenda.  Not only is it a really safe city, but also gave us the opportunity to prove just how efficient these technologies are” concludes Miziuła.

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate

The company joins the prestigious NVIDIA Service Delivery Partner program as one of a handful of preferred partners worldwide providing professional services in deep learning.

deepsense.ai today announced it has joined the NVIDIA Service Delivery Partner Program as a preferred partner in deep learning professional services, ready to play a critical role in helping NVIDIA customers grow and succeed as AI computing spreads across industries.

Tomasz Kułakowski, deepsense.ai CEO, emphasizes the importance of this partnership. “We are very proud to join the select group of companies recommended by NVIDIA as professional deep learning service providers. This ensures that we invest our efforts in the right way and build our competencies around NVIDIA’s cutting-edge AI technology.”

deepsense.ai has experience in delivering both research and commercial data science projects based on machine learning, deep learning and reinforcement learning. The solutions designed by the company include various prediction models for customer analytics, recommendation systems, predictive maintenance, image processing models such as defect detection, product recognition, brand visibility analysis, satellite and drone imagery analytics, as well as reinforcement learning-based vehicle automation and artificial limbs control.

The company also supports organizations in long-term building of data science capabilities in-house by providing them with comprehensive tailored training tracks for technical teams and managers. deepsense.ai is made up of 200 specialists, including over 45 award-winning data scientists, and works with global market leaders in the United States and Latin America, Europe, Asia.

“AI is transforming every industry and having deepsense.ai as a member of our NVIDIA Service Delivery Partner program within the deep learning competency will help customers build their capabilities to take advantage of this opportunity,” said Craig Weinstein, VP of the Americas Partner Organization at NVIDIA.

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate

Researchers from the University of Warsaw, Google AI and deepsense.ai take on a new reinforcement learning challenge on Cloud TPU hardware accelerators. The goal of the experiment is to end-to-end train an artificial intelligence to play video games fully inside a computation graph.

A team from the University of Warsaw, made up of Piotr Miłoś, Błażej Osiński and Henryk Michalewski, has started a collaboration on reinforcement learning research with Łukasz Kaiser from the Google Brain team and with researchers from deepsense.ai. This project is connected to a research program on RL that deepsense.ai started last year.
In the experiment, an artificial intelligence will be end-to-end trained to play video games fully inside a computation graph. Assuming that a game simulator would also be a part of the graph, this could make tasks such as training AI to play video games even faster than what deepsense.ai’s team achieved last year. The intent is to run the training process entirely on Cloud TPUs, which are new machine learning accelerators designed by Google. This will save time previously spent on communication between accelerators and a host computer.

The main experiments are being run on Cloud TPUs via the TensorFlow Research Cloud program and supported by Google Warsaw’s Antonio Gulli, Ignacy Kowalczyk and Maciej Pytel, who are helping us to deploy our experiments on the Google Cloud Platform. TFRC provides ML researchers with access to second-generation Cloud TPUs, each of which provides 180 teraflops of machine learning acceleration.

Henryk Michalewski, a research team leader on the project, offered his appreciation. “Many thanks to Google for sharing early access to Cloud TPUs with us, as well as to Antonio Gulli and Ignacy Kowalczyk for providing the Google Cloud Platform power to deploy our experiments. With such a strong infrastructure, we’re perfectly equipped to tackle our ambitious goal and leverage the research on reinforcement learning efficiency we started last year.”

In 2017, Michalewski’s team from deepsense.ai published a paper describing a new method they had developed to train a robotic arm to grip a can of coke. Their work was recognized by the General Chairs of the Conference on Robot Learning (CoRL) as one of 11 noteworthy papers in the reinforcement learning and robotics category. The team presented the paper in November at Google’s headquarters in Mountain View. The method could be used, for example, to train humanoid robots to combine single steps into a walk or a run.

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate

Tapped to help the UN leverage its competencies in the fast‑evolving field of artificial intelligence, deepsense.ai ran an introductory workshop on deep learning and discussed its practical applications in real‑life use cases.

In December, deepsense.ai trained a team of UN analysts in image recognition, a fast‑growing branch of deep learning. The workshop consisted of a theoretical part and a hands‑on coding session introducing participants to more advanced topics.

Lambert Hogenhout, Chief Analytics, Innovation and Partnerships in the United Nations’ Innovation and Emerging Technologies Department, expressed his satisfaction, saying, “We were very happy with the workshop. The format was good and deepsense.ai’s instructors were excellent as trainers: experts in the subject matter obviously, but also good at conveying the knowledge. The participants very much enjoyed it. Thank you for helping organize this at short notice.”

Hogenhout is tasked with providing the UN and its stakeholders with deeper insights into global macro‑economic and social trends. To do so, he uses leading artificial intelligence and machine learning research with a particular focus on data analytics, machine learning (neural networks), natural language processing, search technologies, bots, translation and cognitive computing.

Marta Piechowiak, Strategy & Business Development Director at deepsense.ai, said “Our custom‑tailored training programs are designed to help organizations build innovative AI solutions in‑house. The first project we did for the UN took place a year ago and we are thrilled to continue the cooperation. We’re also proud to support the goals of such an important international institution as the UN. Such partnerships foster innovation for sustainable development around the world.”

In 2016, deepsense.ai helped the UN recognize the potential of machine learning and deep learning in text analysis for global security tracking and internal information management. The project covered a three‑month proof of concept on monitoring and topic retrieval for global news and detecting propaganda on Twitter. deepsense.ai’s team presented the results at the UN’s first TechNovation briefing to 23 directors and high level officers from several UN Agencies at the UN’s Headquarters in New York.

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate

deepsense.ai’s visual framework for creating Apache Spark applications in a fast, simple and interactive way is going open‑source as of September 26. The company is opening it up to the community and will now focus on further developing its other data science tool, Neptune.

deepsense.ai is making its first product’s code open after two years of development. Seahorse, the data analytics workbench, has become popular since its launch at the Spark Summit in San Francisco in 2015. It’s deepsense.ai’s answer to the increasing interest in fast Big Data processing provided by Apache Spark, an open‑source cluster‑computing framework. Seahorse owes its success to the fresh visual approach it employs to coding, which frees data analysts from knowing Spark internals and makes their work even faster.

deepsense.ai has already published Seahorse code on GitHub. The company will continue to provide support and maintenance for enterprise‑level integrations. Individual users can use the community forum moderated by deepsense.ai engineers to get help with their technical questions.

Piotr Niedźwiedź, deepsense.ai’s CTO and co‑founder, explains, “Seahorse hit 10,000 users this year and the number is still growing. From the beginning, we made the tool highly accessible for everyone, offering it at no cost and easy to download in two versions. Now we’re ready to make it fully open.”

Niedźwiedź’s company will now turn to its primary product, released exactly one year ago at Strata+Hadoop World in New York – Neptune, the Machine Learning Lab. He adds, “Building up Neptune has been in line with our development strategy centered around artificial intelligence. When our data science team started to run tons of machine learning and deep learning experiments for our clients daily, we needed a tool like Neptune to make their work easier and faster. So we created one. Now, Neptune is developing its own community and thanks to the constant feedback from our data scientists it grows with many new features every month.”

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate