Well File:

Results


Topic Modeling with Wasserstein Autoencoders

arXiv.org Artificial Intelligence

We propose a novel neural topic model in the Wasserstein autoencoders (WAE) framework. Unlike existing variational autoencoder based models, we directly enforce Dirichlet prior on the latent document-topic vectors. We exploit the structure of the latent space and apply a suitable kernel in minimizing the Maximum Mean Discrepancy (MMD) to perform distribution matching. We discover that MMD performs much better than the Generative Adversarial Network (GAN) in matching high dimensional Dirichlet distribution. We further discover that incorporating randomness in the encoder output during training leads to significantly more coherent topics. To measure the diversity of the produced topics, we propose a simple topic uniqueness metric. Together with the widely used coherence measure NPMI, we offer a more wholistic evaluation of topic quality. Experiments on several real datasets show that our model produces significantly better topics than existing topic models.


Predicting real-time availability of 200 million grocery items in North American stores

#artificialintelligence

Ever wished there was a way to know if your favorite Ben and Jerry's ice cream flavor is currently available in a grocery store near you? Instacart's machine learning team has built tools to figure that out! Our marketplace's scale lets us build sophisticated prediction models. Our community of over 70,000 personal shoppers scans millions of items per day across 15,000 physical stores and delivers them to the customers. These stores belong to our grocery retail partners like Aldi, Costco, Krogers, Safeway, and Wegmans.


Earth Fare Leverages AI to Optimize Promos

#artificialintelligence

Earth Fare has reported solid improvements in year-on-year top-line sales, a year after adopting artificial-intelligence (AI) technology to help optimize promotions, determining which items to promote and how often to do so. The Asheville, N.C.-based natural and organic grocer, which partnered with Toronto-based AI software company Daisy Intelligence, operates 41 stores and is currently working to grow store units by more than 25 percent annually. Given the competitive landscape, it sought a significant advantage to enable its merchandising team to leverage powerful insights locked within its significant volume of historical sales data. AI-driven insights – which can increase total sales by 3 percent, according to Daisy – helped add significant efficiency and predictability to the sales-planning process, which previously posed challenges. Before deploying the solution, Earth Fare's merchandising and marketing teams had to earmark numerous hours each week for determining which products to promote, and the demands for time to drive innovation for new store activations were challenged as the company began to activate its robust pipeline.


Where's the sugar? Supermarket robot creates product maps as it takes stock

#artificialintelligence

If you think that it's hard to remember where all your favorite products are in the local grocery store, well … it's not just you. In the case of large supermarkets carrying thousands of goods, even employees can have trouble remembering where everything is. That's why Toronto's 4D Retail Technology Corp. developed the stock-taking, store-mapping 4D Space Genius robot. In less than an hour, the self-guiding Segway-based Space Genius can reportedly move along every aisle of an average-sized (43,000 sq ft/3,995 sq m) supermarket or other large store, scanning all of the products and barcodes on display in HD and 3D as it does so. First and foremost, this allows it to create an interactive 3D map of the store, in which the location of every item is indicated.