機械学習の社会実装では、予測精度が高くても、機械学習がブラックボックであるために使うことができないということがよく起きます。
このスライドでは機械学習が不得意な予測結果の根拠を示すために考案されたLIMEの論文を解説します。
Ribeiro, Marco Tulio, Sameer Singh, and Carlos Guestrin. "" Why should i trust you?" Explaining the predictions of any classifier." Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. 2016.
本スライドは、弊社の梅本により弊社内の技術勉強会で使用されたものです。
近年注目を集めるアーキテクチャーである「Transformer」の解説スライドとなっております。
"Arithmer Seminar" is weekly held, where professionals from within and outside our company give lectures on their respective expertise.
The slides are made by the lecturer from outside our company, and shared here with his/her permission.
Arithmer株式会社は東京大学大学院数理科学研究科発の数学の会社です。私達は現代数学を応用して、様々な分野のソリューションに、新しい高度AIシステムを導入しています。AIをいかに上手に使って仕事を効率化するか、そして人々の役に立つ結果を生み出すのか、それを考えるのが私たちの仕事です。
Arithmer began at the University of Tokyo Graduate School of Mathematical Sciences. Today, our research of modern mathematics and AI systems has the capability of providing solutions when dealing with tough complex issues. At Arithmer we believe it is our job to realize the functions of AI through improving work efficiency and producing more useful results for society.
機械学習の社会実装では、予測精度が高くても、機械学習がブラックボックであるために使うことができないということがよく起きます。
このスライドでは機械学習が不得意な予測結果の根拠を示すために考案されたLIMEの論文を解説します。
Ribeiro, Marco Tulio, Sameer Singh, and Carlos Guestrin. "" Why should i trust you?" Explaining the predictions of any classifier." Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. 2016.
本スライドは、弊社の梅本により弊社内の技術勉強会で使用されたものです。
近年注目を集めるアーキテクチャーである「Transformer」の解説スライドとなっております。
"Arithmer Seminar" is weekly held, where professionals from within and outside our company give lectures on their respective expertise.
The slides are made by the lecturer from outside our company, and shared here with his/her permission.
Arithmer株式会社は東京大学大学院数理科学研究科発の数学の会社です。私達は現代数学を応用して、様々な分野のソリューションに、新しい高度AIシステムを導入しています。AIをいかに上手に使って仕事を効率化するか、そして人々の役に立つ結果を生み出すのか、それを考えるのが私たちの仕事です。
Arithmer began at the University of Tokyo Graduate School of Mathematical Sciences. Today, our research of modern mathematics and AI systems has the capability of providing solutions when dealing with tough complex issues. At Arithmer we believe it is our job to realize the functions of AI through improving work efficiency and producing more useful results for society.
AAAI2023「Are Transformers Effective for Time Series Forecasting?」と、HuggingFace「Yes, Transformers are Effective for Time Series Forecasting (+ Autoformer)」の紹介です。
AAAI2023「Are Transformers Effective for Time Series Forecasting?」と、HuggingFace「Yes, Transformers are Effective for Time Series Forecasting (+ Autoformer)」の紹介です。
Takashi Kobayashi and Hironori Washizaki, "SWEBOK Guide and Future of SE Education," First International Symposium on the Future of Software Engineering (FUSE), June 3-6, 2024, Okinawa, Japan
Machine Learning Software Engineering Patterns and Their EngineeringHironori Washizaki
Hironori Washizaki, "Machine Learning Software Engineering Patterns and Their Engineering," 2nd International Workshop on Responsible AI Engineering (RAIE’24), Keynote, Lisbon, April 16th, 2024.
Software Engineering Patterns for Machine Learning ApplicationsHironori Washizaki
Hironori Washizaki, Software Engineering Patterns for Machine Learning Applications, 2021 IEEE International Conference on Electronic Technology, Communication and Information (ICETCI 2021), Keynote, August 28, Online, 2021.
Rubric-based Assessment of Programming Thinking Skills and Comparative Evalua...Hironori Washizaki
Hironori Washizaki, "Rubric-based Assessment of Programming Thinking Skills and Comparative Evaluation of Introductory Programming Environments," 4th International Annual Meeting on STEM Education (IAMSTEM 2021), Keynote, August 12-14, 2021, Keelung, Taiwan and Online
Smart SE: Recurrent Education Program of IoT and AI for BusinessHironori Washizaki
Hironori Washizaki, "Smart SE: Recurrent Education Program of IoT and AI for Business," 2021 IEEE International Conference on Educational Technology (ICET), Keynote, Online, June 20, 2021.
35. eAIおよびML Design Patternsの整理
Topology Programming Model operations
レジリエ
ント
サービ
ング
再現性
責任・説
明性
モデル
訓練
問題
表現
データ
表現
Hashed Feature Embeddings
Feature Cross Multimodal Input
Reframing Multilabel
Ensembles Cascade
Neutral Class Rebalancing
Useful Overfitting
Checkpoints
Transfer Learning
Distribution Strategy
Stateless Serving Function
Hyperparameter
Tuning
Batch Serving Continued Model Evaluation
Keyed Predictions
Windowed Inference
Repeatable Splitting
Transform
Bridged Schema
Two-Phase Predictions
Feature
Store
Model
Versioning
Heuristic Benchmark
Workflow Pipeline
Fairness Lens
Explainable
Predictions
Different Workloads in Different
Computing Environments
Distinguish Business Logic from ML
Models
ML Gateway Routing Architecture
Microservice Architecture for ML
Lambda
Architecture
Kappa
Architecture
Data Lake for ML
Parameter-
Server Abstraction
Data flows
up, Model
flow down
Secure Aggregation
Separation of Concerns
and Modularization of
ML Components
Discard PoC Code
ML Versioning
Encapsulate ML models
within Rule-base Safeguards
Deployable Canary Model
今後の拡充検討エリア