MauroCerbai: udacity

EVALUATION METRICS

The most simple and immediate metric is accuracy

accuracy = labeled correctly / all data

but it depends very much on the number of data in input so with different data is not comparable.

To resolve this we use the confusion matrix

Each row of the matrix represents the instances in a predicted class while each column represents the instances in an actual class (or vice versa).

Analyzing this data we can extract this two data:

recall = how many times you get correctly? (similar to accuracy)

true positive / ( true positive + false negative )

precision = once predicted x, what is the probability that is really x?

true positive / ( true positive + false positive )

No commenti

CROSS VALIDATION

One round of cross-validation involves partitioning a sample of data into complementary subsets, performing the analysis on one subset (called the training set), and validating the analysis on the other subset (called the validation set or testing set).

The conventional validation works partitioning the data set into two sets of 70% for training and 30% for test for example.

Sklearn:

from sklearn import cross_validation

feature_train, feature_test, label_train, label_test = cross_validation.train_test_split (iris_data, iris_target, test_size=0.4, random_state=0)

[train]

pca.fit (feature_train)

pca.transform(feature_train)

svc.train(feature_train)

[test]

NO FIT (you want to use the same function as in the training)

pca.transform(feature_test)

svc.train(feature_test)

K-Fold:

In k-fold cross-validation, the original sample is randomly partitioned into k equal sized subsamples. Of the k subsamples, a single subsample is retained as the validation data for testing the model, and the remaining k − 1 subsamples are used as training data.

So, we can explain also like this:

repat k times

pick 1 block of data as test
train against the othe k-1 block
test on testing set

average final result

Sklearn:

from sklearn.cross_validation import KFold

kf = KFold(len(authors), 2)

for train_indices, test_indices in kf:

feature_train = [word_data[ii] for ii in train_indices]

feature_test = [word_data[ii] for ii in test_indices]

authors_train = [authors[ii] for ii in train_indices]

authors_train = [authors[ii] for ii in test_indices]

GridSearchCV:

Parameter tuning is the process of selecting the values for a model's parameters that maximize the accuracy of the model.

Scikit-learn provides an object that, given data, computes the score during the fit of an estimator on a parameter grid and chooses the parameters to maximize the cross-validation score.

By default, the GridSearchCV's cross validation uses 3-fold KFold or StratifiedKFold depending on the situation.

Sklearn:

parameters = { ‘kernel’: (‘linear’, ‘rbf’), C [1, 10])

svr = svm.SVC

clf = grid_search.GridSearchCV(svr, parameters)

clf.fit(iris_data, iris_target)

print clf.best_params_

No commenti

PRINCIPAL COMPONENT ANALYSIS

PCA find a new coordinates system that is detained from the old one by translation and rotation only centering the data. The goal is to try making a composite feature that more directly probes the underlying phenomenon ( square footage + number of rooms → size ).

How to determine the pca:

The pca of a dataset is the direction that has the largest variance (variance = spread of data distribution) because it retains the maximum amount of original information. That is true because projecting the original data on the longer axis of the new coordinate system we can have a more spread data value and lose the minimum amount of information possible.

When to use:

access latent feature
dimensionality reduction

visualize high dimensional data
reduce noise
use as preprocessing (reducing input for later algo [ eigenfaces] )

Sklearn:

from sklearn.decomposition import PCA

pca = PCA(n_components=2)

pcs.fit(data)

print pca.explained_variance_ratio_

first_pc = pca.components_[0]

second_pc = pca.components_[1]

x_train_pca = pcs.transform(X_test)

No commenti

Every time I open Android Studio this come to mind is this

IMHO

Observer Pattern? MVC? MVP? MVVM? I don't know

Do Androids (Programmer) Dream of Code Order?

Take a look at this post I found useful. Bye

No commenti

I just received this email.

Congratulations!
Dear Mauro,
We are excited to offer you a Google Developer Challenge Scholarship to the Android Developer track.We received applications from many talented and motivated candidates, and yours truly stood out.

I'm very happy to announce that I've been selected for this scholarship involving the famous Google product and the amazing learning platform Udacity. Thank you!

No commenti

OUTLIERS

An outlier is an observation point that is distant from other observations.

Detention:

You simply follow this flow:

Train the algorithm
Remove ~10% of data with the largest residual error
Train again and evaluate the accuracy test

Repeat

The residual of an observed value is the difference between the observed value and the estimated value of the quantity of interest

No commenti

LINEAR REGRESSION

In linear regression, the relationships are modeled using linear predictor functions whose unknown model parameters are estimated from the data. Basically a regression output is not discrete but a function like:

y = ax + b

y is target

a is slope

b is intercept

Error metrics:

The error is calculated

error = actual data - predicted data

The best linear regression is the one that

minimizes all points( actual - predicted)2

algo: ordinary least squares (ols) - gradient descent

But it’s not perfect because it’s an high value if you have multiple data point and lower with fewer, so it’s not comparable very well.

Instead R2

R squared measures the fraction of the variance of the dependent variable expressed by the regression. In simple linear regression it is simply the square of the correlation coefficient. It’s independent from the number of data.

(not good) 0 < r2 > 1 (good)

Sklearn:

from sklearn.linear_model import LinearRegression

clf = LinearRegression()

clf.fit(feature_training, label_training)

prediction = clf.predict(feature_test)

accuracy = clf.score(feature_test, label_test) [ -> R2 error metric ]

slope = clf.coef_

intercept = clf.intercept_

No commenti

NAIVE BAYES

Are a family of simple probabilistic classifiers based on applying Bayes’ theorem with strong independent assumptions between the feature.

Theorem:

P(A|B) = P(B|A) P(A) / P(B)

In a hospital the probability of a liver disease is 10%, the probability of the patient being alcoholic is 5% and among those with a liver disease there are a 7% that are alcoholic. What is the probability of a liver disease if the patient is alcoholic?
P(L) = 0,1

P(A) = 0,05

P(A|L) = 0,07

P(L|A) = P(A|L)*P(L) / P(A) = 0.07*0.01 / 0.05 = 0.14 => 14%

It’s a popular method for text categorization with word frequencies as the features but not their order, it assume that the value of a particular feature is independent of the value of any other feature. Basically it count the occurrences of a word in a particular text sample and assign a probability to that, when you need to attribute a particular “email” to someone then it compare the probability of every word of being written by a certain person.

SENDER : CHRIS - Love 0.1 - Deal 0.8 - Life 0.1

SENDER : SARA - Love 0.5 - Deal 0.2 - Life 0.3

P(CHRIS) = 0.5 = P(SARA)

TEXT : Love deal

P(CHRIS) = 0.1*0.8*0.5 = 0.04 -> 0.04/0.09 = 44%

P(SARA) = 0.5*0.2*0.5 = 0.05 -> 0.05/0.09 = 55%

Sklearn:

from sklearn.naive_bayes import GaussianNB

clf = GaussianNB()

clf.fit(feature_training, label_training)

prediction = clf.predict(feature_test)

accuracy = clf.score(feature_test, label_test)