Lasso RegressionYouTube↗  IA↗(long shot)
Education  22.8K views 112 likes 7 dislikes   23 Feb 2015
Udacity 651K subscribers United States YouTube↗


Please select the problem
You must specify your email and describe the problem in detail
Describe your problem in detail if it's not obvious

Click to Play Video Click to Play Video
So here's the way that a lasso regression can actually make a regression
حسنًا، إليكم الطريقة التي يمكن لانحدار لاسو للتحديد الحر إجراء انحدار
simpler in terms of the number of features that it's using.
.أبسط بها من حيث عدد الميزات التي يستخدمها
So what it does is for features that don't help the regression results enough,
،لذلك، ما يقوم به هو أنه يمكنه تعيين المعامل لهذه الميزات إلى قيمة صغيرة جدًا
it can set the coefficient of those features to a very small value,
،ربما تكون صفرًا
potentially to zero.
.وذلك للميزات التي لا تساعد بشكل كاف في الوصول إلى نتائج الانحدار
So imagine that I have up to four features that are available to me,
،لذلك، تخيلوا أنه يتوفر لدي أربع ميزات، من x1 حتى x4
x1 through x4, and I don't know which ones are the most powerful.
.ولا أعرف أي منها الأقوى
What are the ones that I actually want to be using?
ما الميزات التي أريد استخدامها فعلاً؟
I don't want to use all of these if I don't have to.
.لا أريد استخدامها كلها ما لم أكن مضطرًا لذلك
And that m1 through m4 are the coefficients of regression that I get for
وتعد المعاملات من m1 إلى m4 هي معاملات الانحدار التي أحصل عليها
these features when I actually perform the fit.
.لهذه الميزات عند القيام بالملاءمة
So in an ordinary multivariate regression,
،إذن، في الانحدار متعدد المتغيرات العادي
I would get a result that looks something like this.
.سأحصل على نتيجة تبدو مشابهة لذلك
It will use all of the features that I make available to it and
حيث يتم استخدام جميع الميزات التي أدخلتها
it will assign to each one a coefficient of regression.
.إضافة إلى تعيين معامل انحدار لكل ميزة
What lasso regression will do, is it will try adding them in one at a time.
.ما سيفعله انحدار لاسو للتحديد الحر هو أنه سيحاول إضافتها في وقت واحد
And if the new feature doesn't improve the fit enough to
وإذا كانت الميزة الجديدة لا تعمل على تحسين الملاءمة بشكل كافٍ
outweigh the penalty term of including that feature, then it won't be added.
.لتجاوز مدى الغرامة المترتبة على هذه الميزة، فلن تتم إضافتها
And by not added, it just means that it sets the coefficient to zero.
.وبعدم إضافتها، يعني هذا أنها تعين المعامل إلى صفر
So let's suppose that x1 and
لنفترض أن x1
x2 are two features that are really critical to include in our regression.
،وx2 هما ميزتان مهمتان للغاية يجب إدخالهما في الانحدار
They really describe the pattern with a low of power.
.وهما يصفان بالفعل النمط بقوة أقل
But that x3 and x4 don't improve the fit by very much.
.ولكن هاتين الميزتين x3 وx4 لا تعملان على تحسين الملاءمة بشكل كبير
They might improve it by a little bit, but we could just be fitting to noise.
.يمكن أن تعملا على تحسينها، ولكن يمكن إجراء الملاءمة بتشويش في البيانات
Then what will happen is that the coefficients for
وما سيحدث هو أن المعاملات
those two features will be set to zero.
.لهاتين الميزتين سيتم ضبطهما على صفر
So m3 will be set to 0 and m4 will be set to 0.
.لذلك، سيتم ضبط m3 على 0 وm4 على 0
Which effectively means that these two features become irrelevant in the fit,
،وهو ما يعني عمليًا أن هاتين الميزتين تصبحان غير ذات صلة في الملاءمة
and the only thing that's left is x1 and x2,
،وأن الشيء الوحيد المتبقي هو x1 وx2
the original two features that we wanted to have anyway.
.وهما الميزتان الأصليتان اللتان أردنا وجودهما على أي حال
The power of regularization, the power of having this penalty term for
إن قوة التنظيم وقوة وجود مدى الغرامة
the extra features, is that it can automatically do this selection for you.
.للميزات الإضافية، تكمن في أنه يمكنه القيام بهذا الاختيار لأجلكم
Similar videos:

Introduction Pt. III
Udacity
81.5K views 130 likes 4 dislikes 23 Feb 2015100% match

Bayes Rule
Udacity
28K views 20 likes 1 dislikes 23 Feb 2015100% match

Decision Tree Accuracy - Intro to Machine Learning
Udacity
25.2K views 2 likes 0 dislikes 23 Feb 2015100% match

Kernel Trick
Udacity
64.7K views 161 likes 102 dislikes 23 Feb 2015100% match

Speed Scatterplot 3 - Intro to Machine Learning
Udacity
38.4K views 17 likes 2 dislikes 23 Feb 2015100% match

Nonlinear Data - Intro to Machine Learning
Udacity
37K views 11 likes 5 dislikes 23 Feb 2015100% match

Normalizing 3 - Intro to Machine Learning
Udacity
44.6K views 2 likes 8 dislikes 23 Feb 2015100% match

Accuracy of Naive Bayes - Intro to Machine Learning
Udacity
91.3K views 39 likes 4 dislikes 23 Feb 2015100% match

Why Is Naive Bayes Naive - Intro to Machine Learning
Udacity
64.5K views 58 likes 27 dislikes 23 Feb 2015100% match

Prior and Posterior - Intro to Machine Learning
Udacity
68.8K views 44 likes 19 dislikes 23 Feb 2015100% match

Speed Scatterplot: Grade and Bumpiness - Intro to Machine Learning
Udacity
94.6K views 37 likes 3 dislikes 23 Feb 2015100% match

Playing Around with Kernel Choices - Intro to Machine Learning
Udacity
47.2K views 17 likes 0 dislikes 23 Feb 2015100% match

Practice With Margins - Intro to Machine Learning
Udacity
37K views 5 likes 0 dislikes 23 Feb 2015100% match

Linearly Separable Data - Intro to Machine Learning
Udacity
38K views 30 likes 14 dislikes 23 Feb 2015100% match

Entropy Calculation Part 3 - Intro to Machine Learning
Udacity
28.3K views 2 likes 6 dislikes 23 Feb 2015100% match

Min Samples Split - Intro to Machine Learning
Udacity
25.6K views 4 likes 0 dislikes 23 Feb 2015100% match

Bag of Words - Intro to Machine Learning
Udacity
50.5K views 88 likes 63 dislikes 23 Feb 2015100% match

Information Gain Calculation Part 3 - Intro to Machine Learning
Udacity
26.4K views 3 likes 0 dislikes 23 Feb 2015100% match

Multiple Linear Questions - Intro to Machine Learning
Udacity
31.4K views 10 likes 1 dislikes 23 Feb 2015100% match

Speed Scatterplot 2 - Intro to Machine Learning
Udacity
107.5K views 12 likes 0 dislikes 23 Feb 2015100% match

Gaussian Deployment Solution - Intro to Machine Learning
Udacity
31.9K views 8 likes 2 dislikes 23 Feb 2015100% match

Cancer Test Solution - Intro to Machine Learning
Udacity
24.6K views 8 likes 1 dislikes 23 Feb 2015100% match

GaussianNB Deployment on Terrain Data - Intro to Machine Learning
Udacity
116.6K views 30 likes 1 dislikes 23 Feb 2015100% match

SVM Response to Outliers - Intro to Machine Learning
Udacity
38.4K views 10 likes 2 dislikes 23 Feb 2015100% match

Gaussian NB Example
Udacity
50.3K views 80 likes 19 dislikes 23 Feb 2015100% match

Prerequisites
Udacity
80.5K views 158 likes 10 dislikes 23 Feb 201573% match

Introduction Part II
Udacity
90.9K views 164 likes 12 dislikes 23 Feb 201567% match

Getting Started With sklearn
Udacity
38.4K views 41 likes 2 dislikes 23 Feb 201562% match

NB Decision Boundary in Python
Udacity
38.8K views 17 likes 2 dislikes 23 Feb 201562% match

Features and Labels Musical Example
Udacity
69.5K views 98 likes 8 dislikes 23 Feb 201557% match


Settings Privacy About Contact Donate