SVM Parameters - Practical Machine Learning Tutorial with Python p.33

preview_player
Показать описание

Рекомендации по теме
Комментарии
Автор

topics are more clear due to this lecture, what we can actually do and which one topics are related to another topics ....THank you

clickaws
Автор

Why no practical eg for svm using kernels?

ragvri
Автор

Do you have a video or some reference to a video about image classification with SVM?

MrTravelerX
Автор

Hi, I have 7 independent varibles where 1 needs to be weighted heavier than another 1, but the rest the same. How do I do that? LogisticRegression = (class_weight{4:5}). When I try and fit the classifier to the X and Y training sets it doesnt see the independent varible 4 in the y_train. But y_train has only 1 dependent variable?

Rossboe
Автор

Quick question, these 'from scratch' videos where you make the algos right from scratch are pretty much going from the top of my head. Can I just skip to video 33 (this) since you're gonna start using the actual machine learning methods w/ parameters?

lodashnotebook
Автор

sir, i have extracted the feature of each layer but i dont know how to concatinate into single vector and then how to use in svm

geetanjalisharma
Автор

Hi.. Very good tutorial. I have one question. How can you calculate confusion matrix in this? Do we need to calculate separately ?

kaushikkashyap
Автор

Could you make a video working with the google vision api with face/label detection and just how to use it or even making minor modifications to the code they offer?

lauraomalley
Автор

Hi @sentdex,

a very explanatory tutorial found it very helpful,

I have a question in Input parameters ( X : {array-like, sparse matrix}, shape (n_samples, n_features) ),
how can we add sparse matrix ? considering sparse would be libsvm format ( <label> <index>:<value> <index>:<value>)

thanks

greettheceo
Автор

could you please make a tutorial on One Class SVM, I can't find anyone on Youtube

noureddinehoucineamraoui
Автор

so...how do you classify more than 2 classes in practice?

jrM
Автор

Wich one of those are W and wich ones are B ?

DaLoler
Автор

@sentdex this could be a better way to convert to numeric from non-numeric columns!!!!
import numpy as np
import pandas as pd

df = pd.read_excel("titanic.xls")

# selecting names of columns which are non-numeric
non_numeric = [col for col in df.columns if df[col].dtype!=np.int64 and df[col].dtype!=np.float64]

#column wise finding unique values and replacing it with 1 till the number
# of unique values in the column
for col in non_numeric:
value=df[col].unique()
df[col].replace(dict(zip(value, range(1, len(value)+1))), inplace=True)

prateekgupta
Автор

how to use svm classifier for multi class classification in python ??

geetanjalisharma
Автор

how do you install cvxopt for Python 3.6

sorabhrathod
Автор

Could you give me an example with class_weight as dict?

swarnarampalli
Автор

i like all u tutorial..
keep good tutorial bro..

btw, can i ask, in ovr why 3 vs 8, it's should be 3 vs 7 ?

BudiMulyo
Автор

Can any body give me walkthrough for anomaly detection :fraud detection experiment using one class support vector machine algorithms in azure machine learning .thanks in advance sir. I am very new to machine learning

sultanmdsabir