No abstract

Abstract. E-commerce websites such as Amazon, Alibaba, Flipkart, and Walmart sell billions of products. Machine learning (ML) algorithms involving products are often used to improve the customer experience and increase revenue, e.g., product similarity, recommendation, and price estimation. The products are required to be represented as features before training an ML algorithm. In this paper, we propose an approach called MRNet-Product2Vec for creating generic embeddings of products within an e-commerce ecosystem. We learn a dense and low-dimensional embedding where a diverse set of signals related to a product are explicitly injected into its representation. We train a Discriminative Multi-task Bidirectional Recurrent Neural Network (RNN), where the input is a product title fed through a Bidirectional RNN and at the output, product labels corresponding to fifteen different tasks are predicted. The task set includes several intrinsic characteristics about a product such as price, weight, size, color, popularity, and material. We evaluate the proposed embedding quantitatively and qualitatively. We demonstrate that they are almost as good as sparse and extremely high-dimensional TF-IDF representation in spite of having less than 3% of the TF-IDF dimension. We also use a multimodal autoencoder for comparing products from different language-regions and show preliminary yet promising qualitative results.

Multiple product attributes like dimensions, weight, fragility, liquid content etc. determine the package type used by e-commerce companies to ship products. Sub-optimal package types lead to damaged shipments, incurring huge damage related costs and adversely impacting the company's reputation for safe delivery. Items can be shipped in more protective packages to reduce damage costs, however this increases the shipment costs due to expensive packaging and higher transportation costs. In this work, we propose a multi-stage approach that trades-off between shipment and damage costs for each product, and accurately assigns the optimal package type using a scalable, computationally efficient linear time algorithm. A simple binary search algorithm is presented to find the hyper-parameter that balances between the shipment and damage costs. Our approach when applied to choosing package type for Amazon shipments, leads to significant cost savings of tens of millions of dollars in emerging marketplaces, by decreasing both the overall shipment cost and the number of in-transit damages. Our algorithm is live and deployed in the production system where, package types for more than 130, 000 products have been modified based on the model's recommendation, realizing a reduction in damage rate of 24%.

No abstract

Recent developments in deep learning have shown significant improvement in the accuracy of acoustic impedance inversion results. However, the conventional gradient-based optimizers such as root mean square propagation (RMSProp), momentum, adaptive moment estimation (ADAM), etc., used in the deep learning framework, inherently tend to converge at the nearest optimum point, thereby compromising the solution by not attaining the global minimum. We apply a hybrid global optimizer, genetic-evolutionary ADAM (GADAM) to address the issue of convergence at a local optimum in a semi-supervised deep sequential convolution network-based learning framework to solve the non-convex seismic impedance inversion problem. GADAM combines the advantages of adaptive learning of ADAM and genetic evolution of genetic algorithm (GA), which facilitates faster convergence, and avoids sinking into the local minima. The efficacy of GADAM is tested on synthetic benchmark data and field examples. The results are compared with that obtained from a widely used ADAM optimizer and conventional least-squares method. In addition, uncertainty analysis is performed to check the implication of the optimizer's choice in obtaining efficient and accurate seismic impedance values. Results show that the level of uncertainty and minima of loss function attained using the GADAM optimizer are comparatively lower than that for ADAM. Thus, the present study demonstrates that the hybrid optimizer, i.e., GADAM is more efficient than the extensively used ADAM optimizer in impedance estimation from seismic data in a deep learning framework.