Lying trough, the most amazing paper artifact!

This is a 8 Months ago , The world's top quantitative trading company Optiver stay Kaggle A competition held above —— Forecasting stock market volatility 、 Time series prediction task .

The game attracted more than 3800 There are teams to take part in , Quite a few of them dream of training a model 、 Guide the purchase of shares （OR lottery ）、 Get rich overnight , Marry Bai Fumei and walk to the top of life .

Then of course there is no then .

Leaving aside some of the contestants' dreams that are hard to realize , The popularity of this game represents a phenomenon —— Time series prediction is a traditional technology , Due to the application of machine learning and deep learning methods , New vitality is being radiated .

Generalized linear model ,xgboost And so on ,LSTM,CNN,Transformer Equal depth learning methods can be used for time series prediction

01

ML&DL The background of the rise of time series prediction algorithm

Traditional statistical methods such as ARIMA, ETS, GARCH etc. , Although you don't need a lot of sample data to build a model , However, it requires practitioners to have a deep understanding of time series related statistics , And when there are complex situations such as nonlinear characteristics , Special manual treatment is required , It is not conducive to large-scale prediction .

machine learning & In spite of the poor interpretability of the model in deep learning 、 It needs a large number of sample data to train the model , But this disadvantage is for having a large amount of data 、 In industrial and commercial fields where automation and accuracy are more important than interpretability , It's not a weakness .

02

Industries and scenarios of time series prediction application

01  Financial field

Enterprise cash flow forecast 、 Financial quantification

02 IT field

Intelligent operation and maintenance exception detection, etc

03  Retail pricing

Flight pricing 、 Movie ticket pricing

04  Supply chain

Personnel scheduling plan 、 Production scheduling

  Live broadcast announcement  

6 month 24 Japan -6 month 25 Japan 、 Time series algorithm expert YY, Live sharing ——

01  Introduction to time series prediction （6 month 24 Friday night 20 spot ）

Time series prediction and its application scenarios

Knowledge and skill reserve required for time series prediction

How to analyze the characteristics of time series

02  Introduction to time series prediction algorithm （6 month 25 Friday night 20 spot ）

Prophet Introduction to algorithm principle

Prophet Algorithm code practice

Sweep code payment 0.1 Yuan booking live broadcast

Courseware will be provided after the live broadcast & Code data sets

03

Time series prediction paper list

01  review

Time-series forecasting with deep learning a survey(2021)

02 GNN For time series prediction

Spatio-temporal graph convolutional networks-a deep learning framework for traffic forecasting（2017）

03  Multi level time series prediction

Optimal combination forecasts for hierarchical time series(2011)

Simultaneously Reconciled Quantile Forecasting of Hierarchically Related Time Series(2021)

04 RNN For time series prediction

Deep AR-Probabilistic Forecasting with Autoregressive Recurrent Networks（2017）

Deep and Confident Prediction for Time Series at Uber(2017)

05  Time series prediction in specific fields

Anomaly Detection at Scale-The Case for Deep Distributional Time Series Models（2020）

Reframing demand forecasting- a two-fold approach for lumpy and intermittent demand（2021）

06 CNN For time series prediction

Probabilistic forecasting with temporal convolutional neural network（2020）

A Multi-Horizon Quantile Recurrent Forecaster(2017)

07 ML The algorithm is used for time series prediction

Fast and scalable Gaussian process modeling with applications to astronomical time series（2017）

Forecasting at Scale(2017)

08  The combination of traditional statistics and deep learning

Adjusting for Autocorrelated Errors in Neural Networks for Time Series Regression and Forecasting(2021)

Deep Factors for Forecasting(2019)

Deep State Space Models for Time Series Forecasting(2018)

N-BEATS,Neural basis expansion analysis for interpretable time series forecasting(2020)

09 Transformer For time series prediction

Informer-Beyond Efficient Transformer for Long Sequence Time-Series Forecasting(2020)

Temporal Fusion Transformers for interpretable multi-horizon time series forecasting(2021)

enhancing-the-locality-and-breaking-the-memory-bottleneck-of-transformer-on-time-series-forecasting-Paper(2019)

10  Data enhancement in temporal domain

Improving the Accuracy of Global Forecasting Models using Time Series Data Augmentation（2020）

  Live broadcast announcement  

6 month 24 Japan -6 month 25 Japan 、 Time series algorithm expert YY, Live sharing ——

01  Introduction to time series prediction （6 month 24 Friday night 20 spot ）

Time series prediction and its application scenarios

Knowledge and skill reserve required for time series prediction

How to analyze the characteristics of time series

02  Introduction to time series prediction algorithm （6 month 25 Friday night 20 spot ）

Prophet Introduction to algorithm principle

Prophet Algorithm code practice

Sweep code payment 0.1 Yuan booking live broadcast

Courseware will be provided after the live broadcast & Code data sets