時(shí)空預(yù)測(cè)引領(lǐng)了新的熱點(diǎn)，時(shí)間序列預(yù)測(cè)領(lǐng)域的首個(gè)大模型 TimeGPT 引起業(yè)界熱議，Transformer+時(shí)序，擴(kuò)散模型+時(shí)序更是頂會(huì)新方向大熱“種子”選手，時(shí)序+多方向正在成為這個(gè)AI界矚目的黑馬！

本文整理了時(shí)間序列的時(shí)序預(yù)測(cè) / 時(shí)序-Transformer / 時(shí)序-大模型 / 時(shí)序-擴(kuò)散四大方向的最新論文204篇。

掃碼回復(fù)“時(shí)序”領(lǐng)204篇論文合集

ICLR2024 

ICML 2023

Learning Deep Time-index Models for Time S深度之眼整理eries Forecasting 

KDD 2023

TSMixer: Lightweight MLP-Mixer Model fo深度之眼整理r Multivariate Time Series Forecasting 

因篇幅有限 僅展示前5篇

掃碼回復(fù)“時(shí)序”領(lǐng)204篇論文合集

1.iTransformer: InvertedTransformers Are Effective for Time Series Forecastina

2.Pathformer: Multi- Scale Transformers  With Adaptive Pathways For Time Series Forecasting

3.SCALEFORMER: ITERATIVE MULTI-SCALE REFINING TRANSFORMERS FOR TIME

SERIESFORECASTING

4.InParformer: Evolutionary Decomposition Transformers with Interactive Parallel Attentionfor LongTerm Time Series Forecasting

5.ContiFormer: Continuous-Time Tansformer for Irreqular Time Series Modeling

因篇幅有限 僅展示前5篇

掃碼回復(fù)“時(shí)序”領(lǐng)204篇論文合集

卷積神經(jīng)網(wǎng)絡(luò)方法(4種算法模型)

1.CNN

Recent advances in convolutional neural networks

2.WaveNet-CNN

Conditional time series forecasting with convolutional neural networks

3.Kmeans-CNN

Short-term load forecasting in smart grid: a combined CNN and K-means clustering approach

4.TCN

An empirical evaluation of generic convolutional and recurrent networks for sequence modeling

因篇幅有限 僅展示前4篇

循環(huán)神經(jīng)網(wǎng)絡(luò)方法(3種算法模型）

1.RNN

Bidirectional recurrent neural networks

2.LSTM(長(zhǎng)短期記憶網(wǎng)絡(luò))

Long short-term memory

3.GRU(門控循環(huán)單元)

Learning phrase representations using RNN encoder- decoder for statistical machine translation

Transformer方法(11種算法模型)

1.Transformer

Attention-based models for speech recognition

2.BERT

BERT: pre-training of deep bidirectional transformers for language understanding

3.AST

Adversarial sparse transformer for time series forecasting

4.Informer

Informer: beyond efficient transformer for long sequence time-series forecasting

因篇幅有限 僅展示部分

掃碼回復(fù)“時(shí)序”領(lǐng)204篇論文合集

大模型處時(shí)間序列

1.基于Promtpt的方法

Leveraging Language Foundation Models for HumanMobility Forecasting

2.將時(shí)間序列進(jìn)行離散化處理

AudioLM: a Language Modeling Approach to Audio Generation

3.時(shí)間序列-文本對(duì)齊代表論文

Open Vocabulary Electroencephalography-To-Text Decoding and Zero-shot Sentiment Classification

4.引入視覺(jué)信息

Leveraging Vision-Language Models for Granul深度之眼整理tar Market Change Prediction

5.大模型工具

Unleashing the Power of  Shared Label Structures for Human Activity Recognition

訓(xùn)練時(shí)間序列領(lǐng)域大模型

TimetGPT-1

掃碼回復(fù)“時(shí)序”領(lǐng)204篇論文合集

針對(duì)目前的大熱時(shí)序，我們請(qǐng)來(lái)了頂會(huì)審稿人chichi老師，解讀時(shí)空/時(shí)序預(yù)測(cè)研究現(xiàn)狀與近期熱點(diǎn)。

課程大綱：

- 以GNN為主的時(shí)空預(yù)測(cè)模型 

- 以Transformer為主的時(shí)序預(yù)測(cè)模型 

- LLM在時(shí)空/時(shí)序預(yù)測(cè)上的應(yīng)用

- 針對(duì)時(shí)序新熱點(diǎn)和應(yīng)用總結(jié)

掃碼解鎖

時(shí)序最新熱點(diǎn)解讀直播課