import?numpy?as?np

class?SimpleSVM（object）:
	def?__init__（selftrainxtraintkernel=‘rbf‘kpar?=?1.0C?=?2tol=1e-10max_passes?=?1000）:
		self.trainx?=?trainx
		self.traint?=?traint
		self.kernel?=?kernel
		self.kpar?=?kpar
		self.C?=?C
		self.tol?=?tol
		self.max_passes?=?max_passes
		self.N?=?len（self.traint）
		self.K?=?self.kernel_matrix（）
		self.alpha?=?np.zeros_like（traint）
		self.b?=?0.0

	def?test_kernel（selftestx）:
		testK?=?np.zeros_like（self.traint）
		if?self.kernel?==?‘linear‘:
			testK?=?np.dot（self.trainxtestx.T）[:None]
		elif?self.kernel?==?‘rbf‘:
			for?i?in?range（self.N）:
				testK[i]?=?np.exp（-self.kpar?*?（（self.trainx[i:]-testx）**2）.sum（））
		return?testK

	def?kernel_matrix（self）:
		K?=?None
		if?self.kernel?==?‘linear‘:
			K?=?np.dot（self.trainxself.trainx.T）
		elif?self.kernel?==?‘rbf‘:
			K?=?np.zeros（（self.Nself.N））
			for?i?in?range（self.N）:
				for?j?in?range（self.N）:
					K[ij]?=?np.exp（-self.kpar?*?（（self.trainx[i:]-self.trainx[j:]）**2）.sum（））
		return?K

	def?train_predict（selfi）:
		ksub?=?self.K[i:][:None]
		return?（ksub*self.alpha*self.traint）.sum（）?+?self.b


	def?test_predict（selftestx）:
		testK?=?self.test_kernel（testx）	
		return?（testK*self.alpha*self.traint）.sum（）?+?self.b

	def?smo_optimise（self）:
	????#?initialise
	????self.alpha?=?np.zeros_like（self.traint）
	????self.b?=?0
	????passes?=?0
	????while?passes?	????????num_changed_alphas?=?0
	????????for?i?in?range（self.N）:
	????????????Ei?=?self.train_predict（i）?-?self.traint[i]
	????????????if?（self.traint[i]*Ei??self.tol?and?self.alpha[i]?>?0）:
	????????????????j?=?np.random.randint（self.N）
	????????????????Ej?=?self.train_predict（j）?-?self.traint[j]
	????????????????alphai?=?float（self.alpha[i]）
	????????????????alphaj?=?float（self.alpha[j]）
	????????????????if?self.traint[i]?==?self.traint[j]:
	????????????????????L?=?max（（0alphai+alphaj-self.C））
	????????????????????H?=?min（（self.Calphai+alphaj））
	????????????????else:
	????????????????????L?=?max（（0alphaj-alphai））
	????????????????????H?=?min（（self.Cself.C+alphaj-alphai））
	????????????????if?L==H:
	????????????????????continue

	????????????????eta?=?2*self.K[ij]?-?self.K[ii]?-?self.K[jj]
	????????????????if?eta?>=?0:
	????????????????????continue


	????????????????self.alpha[j]?=?alphaj?-?（self.traint[j]*（Ei-Ej））/eta
	????????????????if?self.alpha[j]?>?H:
	????????????????????self.alpha[j]?=?H
	????????????????if?self.alpha[j]?	????????????????????self.alpha[j]?=?L

	????????????????if?abs（self.alpha[j]-alphaj）?	????????????????????continue

	????????????????self.alpha[i]?=?self.alpha[i]?+?self.traint[i]*self.traint[j]*（alphaj?-?self.alpha[j]）
	????????????????
	????????????????b1?=?self.b?-?Ei?-?self.traint[i]*（self.alpha[i]?-?alphai）*self.K[ii]?-?self.traint[j]*（self.alpha[j]?-?alphaj）*self.K[ij]
	????????????????b2?=?self.b?-?Ej?-?self.traint[i]*（self.alpha[i]?-?alphai）*self.K[ij]?-?s

?屬性????????????大小?????日期????時間???名稱
-----------?---------??----------?-----??----

????.......?????64956??2018-10-18?04:38??機器學習基礎教程源碼\labs\bayesian_regression\.ipynb_checkpoints\Bayesian?Regression?for?the?Olympic?Data-checkpoint.ipynb

????.......????189788??2018-10-18?04:38??機器學習基礎教程源碼\labs\bayesian_regression\bayes_regression2.pdf

????.......???????274??2018-10-18?04:38??機器學習基礎教程源碼\labs\bayesian_regression\data100m.csv

????.......????208218??2018-10-18?04:38??機器學習基礎教程源碼\labs\bayesian_regression\olymp_bayes.pdf

????.......?????27957??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\.ipynb_checkpoints\KNN-checkpoint.ipynb

????.......????160746??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\knn.pdf

????.......????163994??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\logreg.pdf

????.......????170770??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\logreg_metropolis_skeleton.ipynb

????.......????209660??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\naivebayes.pdf

????.......??????3466??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\simplesvm.py

????.......????122629??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\SVM.ipynb

????.......????118909??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\svm.pdf

????.......?????21414??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\testx.csv

????.......??????5357??2018-10-18?04:38??機器學習基礎教程源碼\labs\classification\trainx.csv

????.......????985610??2018-10-18?04:38??機器學習基礎教程源碼\labs\clustering\Gaussian_mixture.ipynb

????.......????146526??2018-10-18?04:38??機器學習基礎教程源碼\labs\clustering\kmeans.pdf

????.......????101392??2018-10-18?04:38??機器學習基礎教程源碼\labs\clustering\K_means_skeleton.ipynb

????.......???????274??2018-10-18?04:38??機器學習基礎教程源碼\labs\linear_regression\data100m.csv

????.......?????42071??2018-10-18?04:38??機器學習基礎教程源碼\labs\linear_regression\linear_regression.ipynb

????.......????144702??2018-10-18?04:38??機器學習基礎教程源碼\labs\linear_regression\linear_regression.pdf

????.......??????2255??2018-10-18?04:38??機器學習基礎教程源碼\labs\linear_regression\max_like.mk

????.......????178518??2018-10-18?04:38??機器學習基礎教程源碼\labs\linear_regression\max_like.pdf

????.......??????2612??2018-10-18?04:38??機器學習基礎教程源碼\labs\linear_regression\vector_matrices_cv.mk

????.......????139534??2018-10-18?04:38??機器學習基礎教程源碼\labs\linear_regression\vector_matrices_cv.pdf

????.......??????1792??2018-10-18?04:38??機器學習基礎教程源碼\matlab\chapter1\.svn\entries

????.......??????2140??2018-10-18?04:38??機器學習基礎教程源碼\matlab\chapter1\.svn\text-base\cv_demo.m.svn-base

????.......??????2140??2018-10-18?04:38??機器學習基礎教程源碼\matlab\chapter1\.svn\text-base\cv_demo.m~.svn-base

????.......???????852??2018-10-18?04:38??機器學習基礎教程源碼\matlab\chapter1\.svn\text-base\fitlinear.m.svn-base

????.......??????1087??2018-10-18?04:38??機器學習基礎教程源碼\matlab\chapter1\.svn\text-base\fitolympic.m.svn-base

????.......??????1021??2018-10-18?04:38??機器學習基礎教程源碼\matlab\chapter1\.svn\text-base\olympicpred.m.svn-base

............此處省略1539個文件信息