YTHAAA的博客

本指南将训练一个神经网络模型，对运动鞋和衬衫等服装图像进行分类。即使您不理解所有细节也没关系；这只是对完整 TensorFlow 程序的快速概述，详细内容会在您实际操作的同时进行介绍。本指南使用了 tf.keras，它是 TensorFlow 中用来构建和训练模型的高级 API。

import numpy as npA = np.array([[1, -1, 3],[-5, 3, 9],[1, 0, -2]])eigenvalue, featurevector = np.linalg.eig(A)print(eigenvalue)print(featurevector)<<<[-3.27491722 1. 4.27491722][[ 0.43401893 0.6882472 -0.33619158][ 0.83410767 0.6

决策树(DTs)是一种用于分类和回归的非参数有监督学习方法。其目标是创建一个模型，通过学习从数据特性中推断出的简单决策规则来预测目标变量的值。决策树的一些优点：1.易于理解和解释。树可以被可视化。2.能够处理多输出问题。3.使用白盒模型。如果给定的情况在模型中是可以观察到的，那么对条件的解释就很容易用布尔逻辑来解释。相反，在黑箱模型中(例如，在人工神经网络中)，结果可能很难解释。4.可以使用统计测试验证模型。这样就有可能对模型的可靠性作出解释。5.即使它的假设在某种程度上被生成数据的真实模型所违

CREATE TABLE website ( id int NOT NULl, name varchar(20) NOT NULL, url varchar(30), age tinyint NOT NULL, alexa int NOT NULL, PRIMARY KEY (id));--创建一个表sp_help website;--查看表的结构

Errorbar limit selectionIllustration of selectively drawing lower and/or upper limit symbols on errorbars using the parameters , of errorbar.uplimslolimsAlternatively, you can use 2xN values to draw errorbars in only one direction.Similarly and can be u

Least-squares minimization (least_squares)The code below implements least-squares estimation of and finally plots the original data and the fitted model function:X

# -*- coding: utf-8 -*-"""Created on Wed Dec 1 23:27:21 2021@author: Machi"""import matplotlib.pyplot as pltimport numpy as npfrom scipy.optimize import curve_fitimport seaborn as snssns.set(color_codes=True)np.random.seed(20211201)e = np.r

import pandas as pdimport numpy as npimport matplotlib.pyplot as pltimport seaborn as sns from statsmodels.graphics.tsaplots import plot_acf from statsmodels.tsa.stattools import adfuller as ADF from statsmodels.graphics.tsaplots import plot_pa

# -*- coding: utf-8 -*-"""Created on Tue Nov 16 20:38:41 2021@author: Machi"""import numpy as npimport matplotlib.pyplot as plt# Fixing random state for reproducibility%config InlineBackend.figure_format = 'retina'np.random.seed(19680801)N =

import numpy as npimport matplotlib.pyplot as plt%config InlineBackend.figure_format = 'retina'np.random.seed(190801)mu = 200sigma = 25n_bins = 50x = np.random.normal(mu, sigma, size=100)fig, ax = plt.subplots(figsize=(8, 4))# plot the cumulat

from scipy.stats import f_onewayimport numpy as npx = np.random.normal(10, 5, 20)y = np.random.normal(10, 5, 20)result = f_oneway(x,y)print(result)F_onewayResult(statistic=0.05215619706951725, pvalue=0.8205774635543236)

# -*- coding: utf-8 -*-"""Created on Mon Nov 15 21:19:53 2021@author: Machi"""import matplotlib.pyplot as pltimport numpy as np%config InlineBackend.figure_format = 'retina'x = np.linspace(0, 2, 100)# Note that even in the OO-style, we use `.py

# -*- coding: utf-8 -*-"""Created on Sun Nov 14 11:09:44 2021@author: Machi"""import numpy as npimport pandas as pdimport seaborn as snsimport matplotlib.pyplot as pltfrom statsmodels.formula.api import olsfrom statsmodels.stats.anova import an

a = rnorm(200)b = rnorm(200)table = data.frame( x = c(a), y = c(b) )aov.manu <- aov(y ~ x, data=table)summary(aov.manu)

# -*- coding: utf-8 -*-"""Created on Sat Nov 13 09:29:27 2021@author: Machi"""import sympy as spx = sp.symbols('x')def f(x): return x**3 - xdef df_value(x,x_value): dy = sp.diff(f(x),x) y_value = float(dy.evalf(

e<-rnorm(100)x1<-filter(e,filter = c(1,-0.5),method = "convolution",circular = T)plot(x1)acf(x1)pacf(x1)

# -*- coding: utf-8 -*-"""Created on Mon Nov 8 21:36:35 2021@author: Machi"""import sympy as spx, y, z = sp.symbols('x y z')func = z * sp.sin(2 * sp.pi ** x + x ** y / 5)func_x = sp.diff(func, x)func_y = sp.diff(func, y)func_z = sp.diff(func,

# -*- coding: utf-8 -*-"""Created on Fri Nov 5 21:44:19 2021@author: Machi"""from scipy.stats import normq = norm.cdf(0,0,1) print(q)print(norm.ppf(q,0,1)) 0.50.0

# -*- coding: utf-8 -*-"""Created on Mon Nov 1 20:20:05 2021@author: 86158"""import numpy as npfrom scipy.stats import normaltestv = np.random.normal(size=100)print(normaltest(v))结果：NormaltestResult(statistic=0.2786470828038562, pvalue=0.86