Search

Jinyun Yan

from San Jose, CA
Age ~41
Get Report
Related to:
Shenao Yan
Connected to:
Adam G Yan, 28Aimin Yan, 68Alicia R Yan, 28Alvin C Yan, 37Anthony K Yan, 54April Yan, 42Betty Yan, 50Brandy Yan, 49Cao Yan, 55Chandy Yan, 40

Jinyun Yan Phones & Addresses

  • San Jose, CA
  • 1030 Hiawatha Ct, Sunnyvale, CA 94087
  • Piscataway, NJ
  • Redmond, WA
  • Edison, NJ
  • Rochester, NY
  • Stanford, CA

Resumes

Resumes

Jinyun Yan Photo 1

Staff Software Engineer

View page
Location:
Palo Alto, CA
Industry:
Computer Software
Work:
Technicolor - Palo Alto, CA Jun 2011 - Dec 2012
Research Intern
Microsoft - Bellevue, WA May 2010 - Aug 2010
Research Intern
Ask.com - Hangzhou, Zhejiang, China Jul 2007 - Aug 2008
Software Engineer
Flexcomm Mar 2006 - Oct 2007
Intern
Education:
Rutgers, The State University of New Jersey-New Brunswick 2008 - 2014
Doctor of Philosophy (Ph.D.), Computer Science Huazhong University of Science and Technology 2005 - 2007
Master's degree, Computer Science
Skills:
Algorithms
Machine Learning
Python
C++
Computer Science
Data Mining
Linux
Matlab
Data Analysis
Java
Interests:
Machine Learning
New Technologies
Hiking
Database
Social Networks
Behavioral Modeling
Startups
Certifications:
Neural Networks and Deep Learning
Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization
Structuring Machine Learning Projects
Convolutional Neural Networks
Sequence Models
Jinyun Yan Photo 2

Jinyun Yan

View page

Publications

Us Patents

Automatically Tuning Parameters In A Layered Model Framework

View page
US Patent:
20210065064, Mar 4, 2021
Filed:
Aug 30, 2019
Appl. No.:
16/557823
Inventors:
- Redmond WA, US
Jinyun Yan - Sunnyvale CA, US
Kinjal Basu - Sunnyvale CA, US
Revant Kumar - Sunnyvale CA, US
Onkar A. Dalal - Santa Clara CA, US
International Classification:
G06N 20/20
G06K 9/62
Abstract:
Techniques are provided for automatically tuning a parameter in a layered model framework. One or more machine learning techniques are used to train multiple versions of a first model that includes a first version and a second version. A second model is stored that includes a parameter and accepts, as input, output from the first model. Multiple parameter values of the parameter are tested when processing content requests using the first and second versions of the first model. A strict subset of the plurality of parameter values are selected for the parameter of the second model, such that processing a first subset of the content requests using the first version of the first model results in a first value of a particular metric that matches a second value of the particular metric resulting from processing a second subset of the content requests using the second version of the first model.

Personalize And Optimize Decision Parameters Using Heterogeneous Effects

View page
US Patent:
20200311745, Oct 1, 2020
Filed:
Mar 29, 2019
Appl. No.:
16/370224
Inventors:
- Redmond CA, US
Kinjal Basu - Sunnyvale CA, US
Jinyun Yan - Sunnyvale CA, US
Shaunak Chatterjee - Sunnyvale CA, US
International Classification:
G06Q 30/02
G06N 5/02
G06N 7/00
Abstract:
Technologies for optimizing content delivery to end-users are provided. Disclosed techniques include storing results of an online experiment with respect to a set of users and determining a plurality of distinct subsets of users based upon the results of the experiment. Users within each of the plurality of distinct subsets may be identified based on metric impacts of the online experiment. For each distinct subset and each associated model parameter, a utility value that represents effectiveness of the model parameter, with respect to an objective, may be determined. An objective optimization model may be used to automatically determine probabilities for each of the model parameters associated with each distinct subset. Users of a second set of users may be assigned to a distinct subset and associated model parameters may be applied to a content delivery strategies of the second set of users.

Position-Aware Corrections In Content Item Selection Events

View page
US Patent:
20200204868, Jun 25, 2020
Filed:
Dec 19, 2018
Appl. No.:
16/224800
Inventors:
- Redmond WA, US
David Pardoe - Mountain View CA, US
Yuan Gao - Sunnyvale CA, US
Jinyun Yan - Sunnyvale CA, US
International Classification:
H04N 21/482
H04N 21/431
H04N 21/45
Abstract:
Techniques for accounting for position-specific differences in user interaction while conducting content item selection events are provided. In one technique, a position-specific factor is determined. The position-specific factor may be based on a ratio of an observed interaction and a predicted interaction. Different positions in a content item feed or on a web page may be associated with different position-specific factors. Eventually, multiple content items are identified for presentation on a screen of a computing device. The content items include a first content item for which a predicted interaction rate is calculated and a second content item for which no predicted interaction rate is calculated. An order of the content items is determined based on the position-specific factor. For example, the predicted interaction rate of the first content item is modified based on the position-specific factor. The content items are presented on the screen based on the order.

Automatically Merging Multiple Content Item Queues

View page
US Patent:
20200099746, Mar 26, 2020
Filed:
Sep 25, 2018
Appl. No.:
16/141669
Inventors:
- Redmond WA, US
Jinyun Yan - Sunnyvale CA, US
Shaunak Chatterjee - Sunnyvale CA, US
Sarah Y. Xing - Sunnyvale CA, US
Gaurav Chandalia - Fremont CA, US
International Classification:
H04L 29/08
Abstract:
Techniques for automatically merging multiple content item queues are provided. In one technique, a first set of content items of a first type is identified. A second set of content items of a second type that is different than first type is identified. The the first set of content items and the second set of content items are merged in a content item feed. Such merging involves, for a particular slot in the content item feed: determining a previous slot that contains a first content item from the first set; determining a number of slots between the previous slot and the particular slot; based on the number of slots, generating a score for a second content item from the second set; and based on the score, determining whether to insert, into the particular slot, the second content item or a third content item from the first set of content items.

Near Real Time Relevance Ranker For Notifications

View page
US Patent:
20190334851, Oct 31, 2019
Filed:
Apr 30, 2018
Appl. No.:
15/966567
Inventors:
- Redmond WA, US
Zhongen Tao - Sunnyvale CA, US
Jinyun Yan - Sunnyvale CA, US
Yan Gao - Sunnyvale CA, US
Shaunak Chatterjee - Sunnyvale CA, US
Sandor Nyako - Sunnyvale CA, US
International Classification:
H04L 12/58
G06F 17/30
H04L 29/08
G06F 15/18
Abstract:
A notification platform for distribution of notification content in an on-line social network system addresses the technical problem of optimizing the volume of quality notifications that are being delivered to a given member. A notification delivery system is designed as a stream processing system that can fetch, store, and process data in a near-line fashion. It can perform feature generation, processing and scoring of notifications, as well as ranking of the notifications based on their respective relevance scores that are calculated using machine learning techniques. The notification delivery system is positioned centrally with respect to different producers of notifications, such that it can consume centrally-stored information about members' holistic notification experiences.

Apparatus And Methods Of Recommending Multiple Options To A Group Of Remote Users And Forming An Agreement

View page
US Patent:
20150339596, Nov 26, 2015
Filed:
Jun 21, 2013
Appl. No.:
14/410206
Inventors:
- Issy de Moulineaux, FR
Jinyun YAN - Sunnyvale CA, US
International Classification:
G06Q 10/02
G06Q 30/06
Abstract:
Methods and apparatus for a system providing group recommendations, event scheduling, and forming consensus agreements among a group of disparately located users are provided by the present principles. Communication among the disparately located users may be achieved through social media tools, online polling, chatting, and texting. The methods and apparatus allow a user profile to be constructed based on ratings that a user provides to items in a database. The items may comprise such things as restaurants, movie theaters, or other entertainment and group activities. The items may comprise feature vectors including attributes of the individual items. With user profiles constructed, the methods and apparatus may allow prediction of a rating that an individual with a particular user profile may give to a similar item in a database. The system may be used to predict the best choice for a group activity by considering the ratings of all users within a group. Individual ratings may be weighted to give higher or lower priority to some users. The system may also recommend activities based on the users participating in the group, or allow the users to select other activities to be considered. In one exemplary embodiment, the system selects recommendations for the users, comprising elements such as movie title, theater location, seating, and prices. The system sends notifications to the users, collects preferences from the users, and determines a choice based on the collected preferences. The users are notified of the choice and given an opportunity to purchase tickets electronically. If so, they may receive electronic tickets, coupons, promotions, or offers.

Method And Apparatus For Contextual Linear Bandits

View page
US Patent:
20150095271, Apr 2, 2015
Filed:
Jun 14, 2013
Appl. No.:
14/402324
Inventors:
- Issy de Moulineaux, FR
Jinyun Yan - Sunnyvale CA, US
Jose Bento Ayres Pereira - Cambridge MA, US
International Classification:
G06N 99/00
G06N 7/00
US Classification:
706 12
Abstract:
A method of selection that maximizes an expected reward in a contextual multi-armed bandit setting gathers rewards from randomly selected items in a database of items, where the items correspond to arms in a contextual multi-armed bandit setting. Initially, an item is selected at random and is transmitted to a user device which generates a reward. The items and resulting rewards are recorded. Subsequently, a context is generated by the user device which causes a learning and selection engine to calculate an estimate for each arm in the specific context, the estimate calculated using the recorded items and resulting rewards. Using the estimate, an item from the database is selected and transferred to the user device. The selected item is chosen to maximize a probability of a reward from the user device.

Method Of Recommending Items To A Group Of Users

View page
US Patent:
20150019469, Jan 15, 2015
Filed:
Dec 13, 2012
Appl. No.:
14/382565
Inventors:
- Issy de Moulineaux, FR
Jinyun Yan - Sunnyvale CA, US
Jose Bento Ayres Pereira - Cambridge MA, US
Assignee:
Thomson Licensing - Issy de Moulineaux
International Classification:
G06N 7/00
US Classification:
706 46
Abstract:
A method for generating a recommendation item for members of a group includes registering a plurality of users as members of the group, identifying a subgroup of members of the group of users wherein the subgroup requests a recommendation item from a recommendation engine, calculating, using a multi-armed bandit algorithm, a recommendation item for the subgroup of members. The recommendation item is provided to the subgroup members for their evaluation. After evaluating the recommendation item, the individual users rate the recommendation item which updates the recommendation engine with preferences representing the members of the subgroup.
Control profile
Jinyun Yan from San Jose, CA, age ~41 Get Report