“I worked with Mr. Wei Zhang and other Almaden researchers developing several Research innovative applications all over a year. During this time he was very flexible to new ideas, and was a great manager dealing with any high level problem, specially when involving solve problems with other teams. It was a pleasure working with him.”
About
Results-driven machine learning and big data expert with 20+ years of experience building…
Activity
-
In the past couple of years, I keep hearing engineers say: I’d like to work on GenAI. At Meta, almost everyone who was thinking about switching teams…
In the past couple of years, I keep hearing engineers say: I’d like to work on GenAI. At Meta, almost everyone who was thinking about switching teams…
Liked by Wei Zhang
-
I'm very excited about the partnership of AMD and Fireworks AI, and the collaboration to launch llama 3.1 models from AI at Meta on MI300X! We will…
I'm very excited about the partnership of AMD and Fireworks AI, and the collaboration to launch llama 3.1 models from AI at Meta on MI300X! We will…
Liked by Wei Zhang
-
It was my 13th #Googleversary ! It's been an amazing journey so far, full of challenges and changes, but I've loved every minute of it.
It was my 13th #Googleversary ! It's been an amazing journey so far, full of challenges and changes, but I've loved every minute of it.
Liked by Wei Zhang
Experience
-
Visa Research
-
-
-
-
-
-
-
-
-
-
Education
Publications
-
Spatial-Temporal Graph Sandwich Transformer for Traffic Flow Forecasting
ECMLPKDD 2023
-
FATA-Trans: Field and Time-Aware Transformer for Sequential Tabular Data
CIKM 2023
-
Spatial-Temporal Graph Boosting Network: Enhancing Spatial-temporal Graph Neural Networks via Gradient Boosting
CIKM 2023
-
An Efficient Content-based Time Series Retrieval System
CIKM 2023
-
Toward a Foundation Model for Time Series Data
CIKM 2023
-
VERB: Visualizing and Interpreting Bias Mitigation Techniques Geometrically for Word Representation
ACM Transactions on Interactive Intelligent Systems(TiiS)
-
Interpretable Debiasing of Vectorized Language Representations with Iterative Orthogonalization
ICLR 2023
-
Learning from Disagreement
Big Data 2023
-
Learning-From-Disagreement: A Model Comparison and Visual Analytics Framework
IEEE Trasactions on Visualization and Computer Graphics(TVCG)
-
Normalization of Language Embeddings for Cross-lingual Alignment
ICLR 2022
-
Error-bounded Approximate Time Series Joins using Compact Dictionary Representations of Time Series
SDM 2022
-
Merchant Identity Recognition Using Credit Card Transactions
IEEE Big Data 2020
-
How Can Self-Attention Networks Recognize Dyck-n Languages?
EMNLP2020
-
Towards Flexible Embedding Learning for Relational Databases, Text, Graphs, etc.
MSDM2020
-
Magellan : A context-aware itinerary recommendation system built only using card transaction data
CIKM2020
-
Multi-stream RNN for Merchant Transaction Prediction
KDD 2020 Workshop on ML in Finance
-
Multi-future Merchant Transaction Prediction
ECML-PKDD 2020
The multivariate time series generated from merchant transaction history can provide critical insights for payment processing companies. The capability of predicting merchants’ future is crucial for fraud detection and recommendation systems. Conventionally, this problem is formulated to predict one multivariate time series under the multi-horizon setting. However, real-world applications often require more than one future trend prediction considering the uncertainties, where more than one…
The multivariate time series generated from merchant transaction history can provide critical insights for payment processing companies. The capability of predicting merchants’ future is crucial for fraud detection and recommendation systems. Conventionally, this problem is formulated to predict one multivariate time series under the multi-horizon setting. However, real-world applications often require more than one future trend prediction considering the uncertainties, where more than one multivariate time series needs to be predicted. This problem is called multi-future prediction. In this work, we combine the two research directions and propose to study this new problem: multi-future, multi-horizon and multivariate time series prediction. This problem is crucial as it has broad use cases in the financial industry to reduce the risk while improving user experience by providing alternative futures. This problem is also challenging as now we not only need to capture the patterns and insights from the past but also train a model that has a strong inference capability to project multiple possible outcomes. To solve this problem, we propose a new model using convolutional neural networks and a simple yet effective encoder-decoder structure to learn the time series pattern from multiple perspectives. We use experiments on real-world merchant transaction data to demonstrate the effectiveness of our proposed model. We also provide extensive discussions on different model design choices in our experimental section.
Other authors -
-
Pcard: Personalized Restaurants Recommendation from Card Payment Transaction Records
theWebConference(WWW) 2019
-
Dynamic Graph Representation Learning via Self-Attention Networks
ICLR 2019 Workshop
-
Pruning Redundant Synthesis Units Based on Static and Delta Unit Appearance Frequency
INTERSPEECH 2015
Other authors -
-
Finding Someone in My Social Directory Whom I Do not Fully Remember or Barely Know
IUI 2012
REACH is an intelligent, people-finding system that helps users to find someone in their social directory, especially those whom they do not fully remember or barely know. It analyzes a user’s communication and social networking data to automatically extract all the contacts and derive multiple facets to characterize each contact in relation to the user. It then employs a personalized, faceted search to retrieve and present a ranked list of matched contacts based on their properties. A…
REACH is an intelligent, people-finding system that helps users to find someone in their social directory, especially those whom they do not fully remember or barely know. It analyzes a user’s communication and social networking data to automatically extract all the contacts and derive multiple facets to characterize each contact in relation to the user. It then employs a personalized, faceted search to retrieve and present a ranked list of matched contacts based on their properties. A preliminary evaluation shows the effectiveness of our approach.
Other authors -
-
Applying Scalable Phonetic Context Similarity in Unit Selection of Concatenative Text-To-Speech
INTERSPEECH 2010
This paper presents an approach using phonetic context similarity as a cost function in unit selection of concatenative Text-to- Speech. The approach measures the degree of similarity between the desired context and the candidate segment under different phonetic contexts. It considers the impact from relatively far contexts when plenty of candidates are available and can take advantage of the data from other symbolically different contexts when the candidates are sparse. Moreover, the cost…
This paper presents an approach using phonetic context similarity as a cost function in unit selection of concatenative Text-to- Speech. The approach measures the degree of similarity between the desired context and the candidate segment under different phonetic contexts. It considers the impact from relatively far contexts when plenty of candidates are available and can take advantage of the data from other symbolically different contexts when the candidates are sparse. Moreover, the cost function also provides an efficient way to prune the search space. Different parameters for modeling, normalization and integerization are discussed. MOS evaluation shows that it can improve the synthesis quality significantly.
Other authorsSee publication -
Recent Improvements of Probability Based Prosody Models for Unit Selection in Concatenative Text-to-Speech
ICASSP 2009
The work presented in this paper is subsequent to the paper “Probability Based Prosody Model for Unit Selection” which was published in ICASSP'2004. In the improved probability prosody model for corpus based concatenative Text-to-Speech (TTS), likelihood is replaced with posterior probability in the cost functions which conduct the following step, unit selection. Objective and subjective experiments show that posterior probability has obvious advantages over likelihood on robustness…
The work presented in this paper is subsequent to the paper “Probability Based Prosody Model for Unit Selection” which was published in ICASSP'2004. In the improved probability prosody model for corpus based concatenative Text-to-Speech (TTS), likelihood is replaced with posterior probability in the cost functions which conduct the following step, unit selection. Objective and subjective experiments show that posterior probability has obvious advantages over likelihood on robustness, flexibility and overall quality.
Other authorsSee publication
Languages
-
English
-
Recommendations received
1 person has recommended Wei
Join now to viewMore activity by Wei
-
An amazing milestone. While I’m not at Cerence anymore, I am proud to have been part of it for a big chunk of these 500M. When we got started on…
An amazing milestone. While I’m not at Cerence anymore, I am proud to have been part of it for a big chunk of these 500M. When we got started on…
Liked by Wei Zhang
-
Oh no, is this for real? With all the hype and panic about AGI being just around the corner, this is what we get? Maybe we should focus more on the…
Oh no, is this for real? With all the hype and panic about AGI being just around the corner, this is what we get? Maybe we should focus more on the…
Liked by Wei Zhang
-
Welcome Mark to join the SIGGRAPH 2024 keynote speakers! DEEPMOTION will present our story to motion technology at the Realtime Live! event of…
Welcome Mark to join the SIGGRAPH 2024 keynote speakers! DEEPMOTION will present our story to motion technology at the Realtime Live! event of…
Liked by Wei Zhang
-
After an incredible 11-year journey with the Siri team, I am retiring from Apple. While I will miss the remarkable team at Apple, I am excited to…
After an incredible 11-year journey with the Siri team, I am retiring from Apple. While I will miss the remarkable team at Apple, I am excited to…
Liked by Wei Zhang
-
In today’s world of AI, data is the key to success. But collecting data isn’t enough. In my latest Salesforce blog, discover how a robust data…
In today’s world of AI, data is the key to success. But collecting data isn’t enough. In my latest Salesforce blog, discover how a robust data…
Liked by Wei Zhang
-
That current LLM architectures won't take us toward something that could be labeled "AGI" ⬇️ Francois Chollet (AI researcher at Google, creator of…
That current LLM architectures won't take us toward something that could be labeled "AGI" ⬇️ Francois Chollet (AI researcher at Google, creator of…
Liked by Wei Zhang
-
Looking forward to welcoming 250 people from the UK and Ireland speech science and speech technology community to Department of Engineering at the…
Looking forward to welcoming 250 people from the UK and Ireland speech science and speech technology community to Department of Engineering at the…
Liked by Wei Zhang
-
In many conversations, I noticed several common misperceptions about generative AI. 1. Technologies behind generative AI are new While many…
In many conversations, I noticed several common misperceptions about generative AI. 1. Technologies behind generative AI are new While many…
Liked by Wei Zhang
-
I'm very excited to share that I have a new Sr Director of Risk Engineering position opened today in my org at Wex to lead a very critical…
I'm very excited to share that I have a new Sr Director of Risk Engineering position opened today in my org at Wex to lead a very critical…
Liked by Wei Zhang
Other similar profiles
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Wei Zhang in United States
1334 others named Wei Zhang in United States are on LinkedIn
See others named Wei Zhang