Recently, Apple researchers, including C. V. Krishnakumar Iyer, Feili Hou, Henry Wang, Yonghong Wang, Kay Oh, Swetava Ganguli, Vipul Pandey, have developed Trinity, a no-code AI platform for complex spatial datasets. 

The platform enables machine learning researchers and non-technical geospatial specialists to experiment with domain-specific signals and datasets to solve various challenges. It tailors complex Spatio-temporal datasets to fit standard deep learning models–in this case, Convolutional Neural Networks (CNNs), and formulate disparate problems in a standard way, eg. semantic segmentation.

READ NEXT

Top 7 Quotes By John McAfee

Recently, Apple researchers, including C. V. Krishnakumar Iyer, Feili Hou, Henry Wang, Yonghong Wang, Kay Oh, Swetava Ganguli, Vipul Pandey, have developed Trinity, a no-code AI platform for complex spatial datasets. 

The platform enables machine learning researchers and non-technical geospatial specialists to experiment with domain-specific signals and datasets to solve various challenges. It tailors complex Spatio-temporal datasets to fit standard deep learning models–in this case, Convolutional Neural Networks (CNNs), and formulate disparate problems in a standard way, eg. semantic segmentation.

Fill the Survey: Utilizing Behavioural Science to Analyze Customer Behaviour

“It creates a shared vocabulary leading to better collaboration among domain experts, machine learning researchers, data scientists, and engineers. Currently, the focus is on semantic segmentation, but it is easily extendable to other techniques such as classification, regression, and instance segmentation,” as per the paper.

Challenges

With the increase in smart devices, a high volume of data containing geo-referenced information is generated and captured. ML techniques have now entered the geospatial domain, including hyperspectral image analysis, high-resolution satellite image interpretation. However, deploying such solutions is still limited due to specific challenges, such as:

• Processing large volumes of Spatio-temporal information and applying ML solutions involves specialised skills and hence has a high barrier of entry, preventing non-technical domain specialists from solving problems on their own.
• The solution differs as data from residential areas will be very different from commercial ones, giving rise to non-standard preprocessing, post-processing, model deployment, and maintenance workflows.
• Engineers process data while scientists run experiments for different problems and involve a lot of back and forth. This hampers the ability to collaborate.

Trinity tackle these challenges by: 

• Bringing information in disparate Spatio-temporal datasets to a standard format by applying complex data transformations upstream. 
• Standardising the technique of solving disparate-looking problems to avoid heterogeneous solutions.
• Providing an easy-to-use code-free environment for rapid experimentation, thereby lowering the bar for entry.

It enables quick prototyping, rapid experimentation and reduces the time to production by standardizing model building and deployment.

The post Now Apple Introduces A No-Code AI Platform appeared first on Artificial Intelligence.

Bengaluru, June 13 (IANS) Amazon India on Sunday announced the launch of ML Summer School which will provide an integrated learning experience for students to gain applied Machine Learning (ML) skills.

A batch of students from select tech campuses in India will be presented with the opportunity to engage through virtual classroom tutorials followed by interactive Q&A sessions with scientists at Amazon.

For students with prior exposure to certain areas of ML, the programme can act as a refresher course, while additionally providing a practical perspective on ML applications in industry, the company said in a statement.

“With the pace of advancements in ML, we are proactively helping students to learn about the latest trends in the field of ML and apply them to solve real-world problems,” said Rajeev Rastogi, VP, India Machine Learning at Amazon.

“Our aim is to prepare students for science roles — this will help to reduce the gap between the growing demand for ML roles across companies and the talent pool with applied ML skills,” he added.

ML Summer School is open to engineering students in their pre-final/final year of Bachelors, Masters or PhD studies across select tech campuses in 2021.

Participants of ML Summer School will be identified through an online assessment.

They will also have access to the Amazon Research Days (ARD) conference where they can learn about technology trends in industry through presentations from renowned ML leaders around the world.

The curriculum of ML Summer School will cover the fundamental concepts in ML while linking them to practical industry applications through an immersive three-day course.

Students will get to learn first-hand on how advanced ML techniques such as Deep Learning and Probabilistic Graphical Models can be leveraged to solve specific business problems in the e-commerce domain such as demand forecasting, catalogue quality, product recommendations, search ranking and online advertising.

The post Amazon India introduces machine learning summer school appeared first on Artificial Intelligence.

Google AI has announced the release of Model Search, a platform that will help researchers develop machine learning (ML) models automatically and efficiently. Model Search isn’t domain-specific, flexible, and well equipped to find the appropriate architecture that best fits a given dataset and problem. At the same time, it minimizes the coding time, effort, and resources. Model Search is built on Tensorflow and can run on both distributed settings or a single machine.

The Success of neural networks often depends on the extent to which they can generalize to various tasks. It is challenging to design Neural networks that can generalize well as the research community’s understanding of this concept is limited. The limitations become complicated when Machine Learning domains are taken into consideration. Techniques like neural architecture search (NAS) use algorithms, reinforcement learning (RL), evolutionary algorithms, and combinatorial search to build a neural network from a given search space. Although these techniques can deliver results better than their manually designed counterparts, these algorithms usually compute heavily and need thousands of models to train before converging and are domain-specific.

These shortcomings can be overcome by using Model Search. The Model Search System is built up of multiple trainers, a search algorithm, and a database to store evaluated models. The system can run both training and evaluation experiments in an adaptive yet asynchronous manner. Each trainer conducts experiments on their own, and all the trainers share knowledge from their experiments. At the starting of every cycle, the search algorithm goes over all the completed trials and then uses beam search to determine what to try next. It then implores mutation over one of the best architectures it finds and assigns the resulting model back to a trainer.

The neural network is built from a set of predefined blocks. This approach is more efficient as it explores only structures and not their fundamental and detailed components, thereby reducing the search space scale. As the framework is built on Tensorflow, blocks can implement any function that takes a tensor as an input. Moreover, the blocks provided can be fully defined neural networks that are already known to work for the given problem. In this case, Model Search can be configured to act as a powerful ensembling machine. The search algorithms used in Model Search are adaptive, greedy, and incremental making them converge faster than RL algorithms.

To improve efficiency and accuracy, Model Search enables transfer learning between various internal experiments in two ways: knowledge distillation or weight sharing. Knowledge distillation allows improving candidates’ accuracy by adding a loss term that matches the high-performing models’ predictions in addition to the ground truth. In contrast, Weight sharing bootstraps some of the network’s parameters from previously trained candidates by copying suitable weights from once trained models and randomly initializing the remaining ones.

The researchers claim that Model Search improves upon production models with minimal iterations. They illustrated Model Search’s capabilities in the speech domain by discovering a model for keyword spotting and language identification. It used fewer than 200 iterations and was found to improve efficiency. The researchers also applied Model Search to find an architecture suitable for image classification on the heavily explored CIFAR-10 imaging dataset. They observed that they were quickly able to reach a benchmark accuracy of 91.83 in only 209 trials as compared to 5807 trials for the RL algorithm.

The Model Search Code aims to provide the researchers with a flexible, domain-agnostic framework for ML model discovery. The framework is powerful enough to build models with state-of-the-art performance on well-known problems when provided with a search space composed of standard building blocks. The code extends access to AutoML solutions to the ever-flourishing research community.

The post Google AI Introduces ‘Model Search’: An Open Source Platform For Finding Optimal Machine learning (ML) Models appeared first on Artificial Intelligence.