趣味でまとめている年表です．参考になれば幸いです．気が向いたときに元論文へのリンクなども追加しています．項目の分け方は著者の趣味です．
大規模言語モデルは別の表にしています．
機械学習，ニューラルネットワークなどの年表
| 年 | 機械学習 | ニューラルネットワーク | その他 |
|---|---|---|---|
| 1700s | ベイズ定理(Bayes) | ||
| 1795 | 最小二乗法 (Gauss) | ||
| 1865 | Clausius' entropy (Clausius) | ||
| 1870s | Boltzmann's entropy (Boltzmann) | ||
| 1873 | ゴルジ染色 (Golgi) | ||
| 1888 | Neuron doctrine (Cajal) | ||
| 1901 | PCA (Pearson) | ||
| 1905 | Random walk (Pearson) | ブラウン運動 (Einstein) | |
| 1906 | Receptive field (Sherrington) | ||
| 1912-22 | 最尤推定 (Fisher) | ||
| 1920 | Ising model (Lenz) | ||
| 1938 | Receptive field (Hartline) | ||
| Turing's BombeがEnigma解読 | |||
| 1943 | Neural network (Neuron) model (McCulloch and Pitts) | ||
| 1946 | ENIAC完成・使用開始 | ||
| 1948 | Shannon's entropy (Shannon) | Jeffress model (Jeffress) | |
| 1949 | Hebbian Learning (Hebb) | ||
| 1950s | Checker (Samuel) | ||
| 1950 | Turing test | ||
| 1951 | SNARC迷路を解く (Minsky and Edmons) | ||
| 1952 | Hodgkin Huxley model (Hodgkin and Huxley) | ||
| 1956 | Artificial Intelligence (Dartmouth meeting) | ||
| 1957 | Bellman equation (Bellman) | Perceptron (Rosenblatt) | |
| 1957 | k-means published in 1980 (Lloyd) | ||
| 1959 | Machine learning (Samuel) | V1 (Hubel and Wiesel) | |
| 1960 | Delta rule (Widrow and Hoff) | ||
| 1966 | Hidden Markov Model (Baum and Petrie) | Chatbot Eliza (Weizenbaum) | |
| 1967 | (Online) k-means (MacQueen) | ||
| 1969 | Perceptrons (Minski and Papert) | ||
| Rectified linear (Fukushima) | |||
| 1971 | a deep GMDH network with 8 layers (Ivakhnenko) | ||
| 1973 | Self organizing map (SOM) (von der Malsberg) | ||
| 1977 | EM Algorithm (Dempster et al.) | ||
| 1980 | Neocognitron (Fukushima) | PacMan | |
| 1980 | SOM (Amari) | MS-DOS | |
| 1982 | Hopfield Network (Hopfield) | ||
| Vision (Marr) | |||
| 1983 | SOM (Kohonen) | ||
| 1986 | Backpropagation (Rumelhart, Hinton, Williams) | ||
| Boltzmann Machine (Ackley, Hinton, Seinowski) | |||
| 1988 | Autoencoder (Baldi and Hornik) | ||
| Autoencoder (Bourland and Lecun) | |||
| 1989 | LeNet (LeCun) | ||
| 1989 | Neurogammon (Tesauro) | ||
| 1994 | ICA (Pierre) | ||
| 1995 | Positive Matrix Factorization (Paatero) | TD-gammon (Tesauro) | |
| SVM (Vapnik) | |||
| 1996 | |||
| Sparse Coding (Olshausen) | |||
| 1997 | LSTM (Hochreiter and Schmidhuber) | Deep Blue beat Kasparov. | |
| Logistello beat Murakami. | |||
| 1998 | Kernel PCA (Scholkopf et al.) | ||
| Quantum annealing | |||
| 1999 | Non-Negative Factorization (Lee) | Aibo (Sony) | |
| 2001 | Topographic ICA (Hyvarinen et al.) | Echo state network (Jaeger) | |
| 2002 | Liquid state machine (Maass) | Roomba (iRobot) | |
| 2003 | Izhikevich model (Izhikevich) | ||
| 2006 | IVA (Kim et al.) | Deep Belief Network (Hinton and Salakhutdinov) | |
| Monte Calro Tree Search (Coulom) | |||
| 2007 | Stacked Autoencoder (Bengio et al.) | ||
| 2010 | Rectufied linear (Nair et al.) | ||
| 2012 | AlexNet (Krizhevsky, Sutskever, Hinton) | ||
| Dropout (Hinton et al.) | |||
| 2013 | Variational Autoencoder (Kingma and Welling) | ||
| 2014 | GoogLeNet (Szegedy et al.) | ||
| VGG (Simonyan and Zisserman) | |||
| 2015 | ResNet (He et al.) | ||
| Deep Q network (Mnih et al.) | |||
| 2016 | Xception (Chollet) | ||
| AlphaGo (Silver et al.) | |||
| 2017 | Squeeze-and-Excitation Networks (Hu et al.) | ||
| 2017 | Transformer (Vaswani et al.) | ||
| AlphaGo Zero (Silver et al.) | |||
| 2018 | AlphaZero (Silver et al.) | ||
| R2D2 | |||
| BERT | |||
| AlphaFold 1 | |||
| 2019 | EfficientNet (Tan and Le) | ||
| AlphaStar | |||
| MuZero (Schrittwieser et al.) | |||
| 2020 | Vision Transformer | ||
| Diffusion model (Ho et al.) | |||
| 2021 | EfficientNetV2 (Tan and Le) | ||
| MLP-Mixer | NovelAI (Anlatan) | ||
| Latent Diffusion | |||
| 2022 | Gato | DALL-E 2 (OpenAI) | |
| Midjourney (Midjourney) | |||
| Stable Diffusion | |||
| ChatGPT (OpenAI) | |||
| 2023 | RT2 (Google) | ||
| Gemini (Google) | |||
| 2024 | Copilotキー搭載を発表 (Microsoft) | ||
| Apple Intelligenceを発表 (Apple) | |||
大規模言語モデル
| 年月 | LLM |
|---|---|
| 2018/6 | GPT (OpenAI) |
| 2018/10 | BERT (Google) |
| 2019/2 | GPT-2 (OpenAI) |
| 2020/5 | GPT-3 (OpenAI) |
| 2021/12 | Claude (Anthropic) |
| 2022/1 | LaMDA (Google) |
| 2022/4 | PaLM (Google) |
| 2022/3 | GPT-3.5 (OpenAI) |
| 2023/2 | LLaMA (Meta) |
| 2023/3 | GPT-4 (OpenAI) |
| 2023/5 | PaLM (Google) |
| 2023/7 | LLama2 (Meta) |
| Claude2 (Anthropic) | |
| 2023/11 | Claude 2.1 (Anthropic) |
| Grok-1 (x.AI) | |
| 2023/12 | Gemini 1.0 (Google) |
| Mixtral 8x7B (Mistral AI) | |
| Phi-2 (Microsof) | |
| 2024/2 | Gemini 1.5 (Google) |
| Gemma (Google) | |
| 2024/3 | Claude3 (Anthropic) |
| 2024/4 | LLama3 (Meta) |
| Mixtral 8x22B (Mistral AI) | |
| Phi3 (Microsof) | |
| 2024/5 | Deepseek V2 (DeepSeek) |
| 2024/6 | Qwen2 (Alibaba) |
| Nemotron-4 (NVIDIA) | |
| 2024/7 | LLama3.1 (Meta) |
| Gemma2 (Google) |