機器學習
觀
機器學習(英文machine learning)者,屬人工智能,蓋計算機科學所鑽研也。其為術也,立演算之法,設數之模,使算器不假明令,[一]積驗而益精其事[二]。是學與統計、臻善(華言:最佳化)、資訊論、控制論相表裏,今廣施於辨像、識音、解文、薦物、診疾(華言:診斷)、御車諸事。[三]
是學之本,在以數模格物,藉演算調參,俾預測決策之誤漸消。今有入變、出變,欲求映射,使遇新而可推其。乃立損失函數以度所推與實之差,復求其極小。[四]術曰:置初參,求損失隨參之變率,以步長乘之,以減其參。損之又損,以至於不可復損,則模成矣。
西元一九五〇年,圖靈著論,問「算器能思乎」,以為當以習代規,使器顯智。[五]一九五七年,羅森布拉特創感知機,實神經網絡之先聲。[六]越二年,薩繆爾始命之名曰「機器學習」。[七]
八十年代以降,決策樹、支援向量機、貝葉斯法諸術並興。其間,一九八六年,魯梅爾哈特、辛頓等倡反向傳播之法,多層神經網絡由是可訓。[八]及二十一世紀,算力日增,與數日夥,深度神經網絡遂破舊障,深度學習由是為顯學。[九]
- ↑ Hügle, Maria; Omoumi, Patrick; van Laar, Jacob M.; Boedecker, Joschka; Hügle, Thomas. Applied machine learning and artificial intelligence in rheumatology. Rheumatology Advances in Practice. 2020, 4 (1): rkaa005. doi:10.1093/rap/rkaa005.
- ↑ Mitchell氏(名T. M.)《Machine Learning》第二頁。McGraw-Hill以一九九七年刊諸New York。
- ↑ Jordan, M. I.; Mitchell, T. M. Machine learning: Trends, perspectives, and prospects. Science. 2015, 349 (6245): 255–260.
- ↑ Hastie氏(名T.)、Tibshirani氏(名R.)、Friedman氏(名J.)《The Elements of Statistical Learning》2nd。Springer以二〇〇九年刊之。
- ↑ Turing, A. M. Computing Machinery and Intelligence. Mind. 1950, 59 (236): 433–460.
- ↑ Rosenblatt, F. The Perceptron: A Probabilistic Model for Information Storage and Organization in the Brain. Psychological Review. 1958, 65 (6): 386–408.
- ↑ Samuel, A. L. Some Studies in Machine Learning Using the Game of Checkers. IBM Journal of Research and Development. 1959, 3 (3): 210–229.
- ↑ Rumelhart, D. E.; Hinton, G. E.; Williams, R. J. Learning representations by back-propagating errors. Nature. 1986, 323: 533–536.
- ↑ LeCun, Y.; Bengio, Y.; Hinton, G. Deep learning. Nature. 2015, 521: 436–444.