Image 04 February 2019
Image
 
Image

Banking

AI To Screen Loan Applications for Regional Japanese Banks

Japanese internet bank, SBI Sumishin Net Bank will work with Hitachi to offer an instant loan-screening service powered by AI to regional banks, a sector facing intensifying competition amid the country's shrinking population. The AI will take only seconds to compute risks and express them as a value between zero and 100.0%, using data such as annual income, occupation, family composition and credit history. Bank staffers conventionally place applicants into several different risk levels based on credit history, a process that takes days. The plan is to start out with mortgages, aiming to serve tens of the nation's 100 or so regional banks within two to three years. The partners intend to broaden to other areas such as card loans and lending to small and mid-size businesses. The AI will pool data from different regional banks for increased accuracy. It will update information provided by the banks every six months to a year for reassessment. Read More

Image
Image

Transport

AI Helps Detect Virtual Blind Spots in Self-Driving Cars

Microsoft and MIT have developed a model that can detect virtual blind spots in self-driving cars. The approach has the AI compare a human's actions in a given situation to what it would have done, and alters its behavior based on how closely it matches the response. If an autonomous car doesn't know how to pull over when an ambulance is racing down the road, it could learn by watching a flesh-and-bone driver moving to the side of the road. The model would also work with real-time corrections. If the AI stepped out of line, a human driver could take over and indicate that something went wrong. Researchers even have a way to prevent the driverless vehicle from becoming overconfident and marking all instances of a given response as safe. This technology isn't ready for the field yet. Scientists have only tested their model with video games, where there are limited parameters and relatively ideal conditions. Read More

Image
Image

Miscellaneous

MIT’s New Robot Uses AI to Play Jenga

Teaching a robot how to play Jenga is a lot more difficult than it sounds. Rather than relying on visual information alone, players have to poke, tap, and feel individual wooden blocks to choose which one to remove from the tower. But thanks to machine learning algorithms, MIT researchers were able to teach a robot how to successfully play Jenga by only giving it a basic set of instructions. The research paper, published by journal Science Robotics today, describes how the robot takes a thorough look at the tower to examine the state of each block. Then it figures out its next move for “successful extraction” of pieces by predicting a block’s future state. It can either push or pull a piece, a single millimeter at a time. Force sensors help continually analyze the situation to figure out if something is wrong or if a tower collapse is imminent. The robot can learn from past mistakes, adjusting its behavior after the tower collapses by “building nuggets of experience. Read More

Image
Image

Tourism

AI to Direct Foreign Tourists on Journeys in Japan

Fujitsu Limited has developed a system that uses AI to recommend travel destinations to foreign tourists. A trial run of the system will begin as early as this month, with Fujitsu aiming for practical application in fiscal 2019. The trial will be jointly conducted with EXest, a company based in Shibuya Ward, Tokyo, that operates WOW U, an English language website that introduces guide services to foreign visitors. Those using the AI system during the trial will provide answers on seven points, such as their age and the goal of their travel, via the WOW U website. The system will then suggest multiple options from among 280 sightseeing spots from Hokkaido to Okinawa. The AI will base its analysis on data collected ahead of time by survey from about 1,000 foreign nationals, and the opinions of guides and other nationals who are registered on the WOW U website. Read More

Image
Image

Healthcare

AI Fills the Gaps in a Patient’s Medical Data

MIT researchers have developed a model that can assimilate multiple types of a patient’s health data to help doctors make decisions with incomplete information. The MIT researchers describe a single neural network that takes as input both simple and highly complex data. Using the known variables, the network can then fill in all the missing variables. For example, looking at the data from a patient’s electrocardiography (ECG) signal, which measures heart function, and self-reported fatigue level, the model can predict a patient’s pain level, which the patient might not remember or report correctly. The network works by stitching together various sub-models, each tailored to describe a specific relation among variables. The sub-models share data as they make predictions, and ultimately output a predicted target variable. The network could also help quantify sometimes-ambiguous health variables for patients and doctors, such as pain and fatigue levels. Read More

Image
Image

Healthcare

AI-based Algorithm Learns to Detect Tumours in Microscopy Images

By applying deep-learning techniques to a set of phase-contrast microscopy images, Japanese researchers from Osaka University have been able to identify the nature and origin of different cancer cells with 96.0% accuracy. This approach could lead to better cancer treatments. The researchers used a convolutional neural network (CNN), a common scheme used in deep learning, to analyse the images. CNNs work by applying to the input image a set of connected filters and mathematical functions that, similarly to neurons, can be trained to extract specific features. In medical imaging, CNNs are modelled on the human visual system, with low layers that capture fine details such as edges, and higher levels that capture complex features reflecting the whole image. In this study, the researchers designed the CNN to classify cells into five categories: untreated (control), X-ray-resistant, carbon-ion beam-resistant mouse tumours, untreated and X-ray-resistant human cervical tumours. They subsequently trained the CNN with a database of 8000 phase-contrast microscopic images containing these types of cells and validated it by using an additional 2000 images. Read More

Image
Image

Apiculture

AI to Save Bees

A beekeeper teamed up with the Signal Processing Laboratory 5 and a group of EPFL students to develop an app that counts the number of Varroa mites in beehives. This parasite is the leading cause of bee deaths. Beekeepers currently assess infestations by counting the number of dead mites that land on a wooden board placed below the hives. But this technique is not very accurate and is also time-consuming. The students came up with a system – consisting of an app linked to a web platform – that uses AI to quickly and automatically count the mites on the boards. The beekeepers still need to put wooden boards under each of their hives, but now they simply photograph the boards and upload the images to the web platform. To develop their app, the students used machine learning – scanning thousands of images into a computer – to teach their program how to recognize the mites. The app can spot and count the dead parasites on the board in just seconds. Read More

Image
Image

Decimal Point Analytics Pvt Ltd does not make any recommendation, solicitation, or offer for any securities and is not responsible for suitability of any securities for any purpose, investment or otherwise. It is the sole responsibility of the client, as a professional organization, to exercise professional due diligence in ensuring suitability of investment and ensuring that when the client publishes a part or full report under its own brand, the legal requirements for distribution of such material are complied with in all the jurisdiction in which it is published. Decimal Point Analytics Pvt. Ltd. shall not be responsible for any loss suffered by the user. The returns indicated, including future projections, in any investment report prepared by Decimal Point Analytics Pvt Ltd are not guaranteed in any manner and may not be achieved.

India :

5A, B-Wing, Trade Star Building, J. B. Nagar, Andheri-Kurla Road, Andheri (East), Mumbai - 400 059, Maharashtra, India
+91 22 30015200
info@decimalpointanalytics.com

United Kingdom:

1st Floor, 99, Bishopsgate, London, EC2M 3XD, United Kingdom

+ 44 20 3286 1998
info@decimalpointanalytics.com

USA:

17 State Street, Suite 4000, New York, NY 10004 U.S.A

+1 (917) 341 3218
info@decimalpointanalytics.com

www.decimalpointanalytics.com