2021-06-24 읽을거리

python

1. 간단하지만 유용한 pandas 사용 방법 3가지

https://towardsdatascience.com/3-pandas-functions-that-will-make-your-life-easier-4d0ce57775a1

3 Pandas Functions That Will Make Your Life Easier

 

2. 유용한 built-in 함수 10가지

https://medium.com/pythoneers/10-must-known-built-in-functions-in-python-2f196b9c0359

10 Must Known Built-In Functions in Python

 

3. python의 loop 문 잘 쓰는 방법

https://medium.com/techtofreedom/the-art-of-writing-loops-in-python-68e9869e4ed4

The Art of Writing Loops in Python

 

4. 잘 모르지만 알아두면 편한 python function

https://levelup.gitconnected.com/python-features-you-probably-dont-know-about-but-should-a66c6b30c528

Python Features You Probably Don’t Know About, But Should

 

---

 

UI/UX, prototyping

1. gradio

UI 컴포넌트를 간단하게 보여줄 수 있는 라이브러리

DL/ML 에 붙여서 사용하기 좋아보임

https://github.com/gradio-app/gradio

gradio-app/gradio

 

---

 

Data Science

1. EDA 하는 방법

https://medium.com/analytics-vidhya/how-to-ace-exploratory-data-analysis-d3821011532b

How to ace Exploratory Data Analysis

 

2. kubeflow에 대한 introduction

https://medium.com/@bv_subhash/kubeflow-1-0-quick-overview-d515834d3c67

Kubeflow 1.0 — Quick Overview

 

3. 유용한 data science 툴들

https://yanwei-liu.medium.com/some-useful-data-science-tool-i-learned-from-full-stack-deep-learning-21c07db7b20c

Some useful Data Science tool I learned from Full Stack Deep Learning

 

4. 수식없이 deep learning의 네트워크 6가지의 특징을 간단하게 설명

https://medium.com/cloud-believers/quickly-understand-the-6-types-of-neural-networks-5e5300c45701

Quickly Understand the 6 Types of Neural Networks

 

5. AutoML Framework들을 비교한 글

https://medium.com/georgian-impact-blog/choosing-the-best-automl-framework-4f2a90cb1826

Choosing the best AutoML Framework

 

6. Tensorflow Lite를 사용한 Object Detection에 대한 코드 예제

https://blog.tensorflow.org/2021/06/easier-object-detection-on-mobile-with-tf-lite.html

Easier object detection on mobile with TensorFlow Lite

 

7. scikit-learn의 pipeline을 사용하여 model tuning, optimization을 하는 코드 예제

https://machinelearningmastery.com/modeling-pipeline-optimization-with-scikit-learn/

Modeling Pipeline Optimization With scikit-learn

 

8. facebook에서 만든 다양한 종류의 데이터들에 대한 Augmentation을 수행하는 라이브러리

README 아래쪽에 예제 링크가 존재

https://github.com/facebookresearch/AugLy

facebookresearch/AugLy

 

9. XAI 알고리즘 : LIME 알고리즘에 대한 설명

https://myeonghak.github.io/xai/XAI-LIME(Local-Interpretable-Model-agnostic-Explanation)-알고리즘/ 

[XAI] LIME(Local Interpretable Model-agnostic Explanation) 알고리즘

---

 

프로그래밍

1. 시스템 디자인 패턴 5가지에 대한 비교

https://betterprogramming.pub/top-5-distributed-system-design-patterns-ae9482f49128

Top 5 Distributed System Design Patterns

 

---

 

DL/ML 모델, 논문

1. XBNet : tabular data 도메인에서 neural network가 사용된 tree-based model

boosted gradient descent 사용

https://github.com/tusharsarkar3/XBNet

tusharsarkar3/XBNet

 

2. Decision Transformer

https://arxiv.org/abs/2106.01345v1

Decision Transformer: Reinforcement Learning via Sequence Modeling

 

3. XCiT

https://arxiv.org/abs/2106.09681v1

XCiT: Cross-Covariance Image Transformers

 

4. A survey of Transformers

현재까지 나온 Transformer 모델들에 대해 분류, 정리한 논문

https://arxiv.org/pdf/2106.04554.pdf