San Francisco, California, United States
Contact Info
8K followers
500+ connections
About
Articles by Roman
Contributions
-
You're struggling to create an algorithm for a recommendation engine. What are some ways to improve it?
Understanding the data is crucial for creating a better relevance algorithm. Data processing teaches machine learning and AI algorithms to work as intended, and the relevance of the data is essential for the algorithms to learn and make accurate decisions. In the field of data science, the importance of algorithms and data structures lies in their efficiency, which is vital for processing and analyzing large volumes of data. Identifying and removing irrelevant data is also a critical step in improving data mining results, as too much irrelevant data can confuse the model and affect its accuracy. Therefore, understanding the data is fundamental to developing an effective relevance algorithm that can provide accurate and meaningful results.
Activity
-
We’re proud to announce that Seekr is now an “Awardable” vendor to the United States Department of Defense via the Tradewinds Solutions Marketplace…
We’re proud to announce that Seekr is now an “Awardable” vendor to the United States Department of Defense via the Tradewinds Solutions Marketplace…
Liked by Roman K.
-
Learn from my idiocy :). My tinkering with SearchArray+BEIR with many humbling lessons about how real search engines work compared to a tool that…
Learn from my idiocy :). My tinkering with SearchArray+BEIR with many humbling lessons about how real search engines work compared to a tool that…
Liked by Roman K.
-
and after some training Meta's LLM gave a proper answer...
and after some training Meta's LLM gave a proper answer...
Shared by Roman K.
Experience & Education
Licenses & Certifications
Volunteer Experience
-
Fundraising
Chabad
- Present 28 years 8 months
Social Services
Raising large capital to sustain Chabad organizations.
Courses
-
Advanced C++
CIS 542
-
Akka
http://lightbend.com/
-
Cassandra
DataStax
-
Cloud Computing
Lynda.com
-
Functional Programming Principles in Scala
Coursera
-
Go: Design Patterns for Real-World Projects
-
-
Google Cloud Platform for Machine Learning Essential Training
-
-
Kotlin From the Ground Up
-
-
Machine Learning
AI Frontiers
-
Microsoft Azure Exam 70-535 Certification Course
-
-
Scala
Fédérale de Lausanne
-
Scala Advanced topics
http://scalacourses.com
-
Software Engineering
CIS 375
-
Solr Under the Hood
Lucidworks University
-
Solr Unleashed
Lucidworks University
Projects
-
https://github.com/RusticRoman/CSharper
C# code for coding assignments
-
Fixed bugs in Lucene 6.4 and submitted to Apache Foundation
Found number of exceptions coming from Solr and decided to contribute back to community by forking the open source project and fixing the bugs.
-
algo4
-
Port of Algorithms for Java into Rust
Languages
-
Java
Full professional proficiency
-
Scala
Full professional proficiency
-
Python
Full professional proficiency
-
C++
Professional working proficiency
-
Kotlin
Professional working proficiency
-
Golang
Professional working proficiency
Organizations
-
ACM
-
- Present
Recommendations received
18 people have recommended Roman
Join now to viewMore activity by Roman
-
LLMs are alien beasts. It is deeply troubling that our frontier models can both achieve silver medal in Math Olympiad but also fail to answer "which…
LLMs are alien beasts. It is deeply troubling that our frontier models can both achieve silver medal in Math Olympiad but also fail to answer "which…
Liked by Roman K.
-
Meta's LLAMA 3.1 hallucinating and not "changing it's mind", even when I hint to it. So does 9.9 or 9.11 greater? I'll go and ask my 7 years old…
Meta's LLAMA 3.1 hallucinating and not "changing it's mind", even when I hint to it. So does 9.9 or 9.11 greater? I'll go and ask my 7 years old…
Shared by Roman K.
-
Apple presents LazyLLM Dynamic Token Pruning for Efficient Long Context LLM Inference paper page: https://buff.ly/4d9En9A The inference of…
Apple presents LazyLLM Dynamic Token Pruning for Efficient Long Context LLM Inference paper page: https://buff.ly/4d9En9A The inference of…
Liked by Roman K.
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 Roman K. in United States
-
Roman K.
-
Roman K
Bioresource Science and Engineering at University of Washington
-
Roman k
Ektech is a one stop, standard and licensed call center that is fully equipped with the latest technology in the telemarketing and call center industry. We are specialized in CGM,DME and many more.
-
Roman K.
-
Roman K
Systems Engineer and Program Manager
23 others named Roman K. in United States are on LinkedIn
See others named Roman K.