Mariusz Gil "Machine Learning"

Mariusz Gil
Source Ministry
@mariuszgil
Machine Learning to a Rescue

IF-OLOGY
UGLY CODE
FOR POC
1STAPPROACH

NAIVE
MACHINE LEARNING
2ND APPROACH

DATA ML
TASK
SENDTO RESULTS
CALCULATE

RECIPE FOR A FAILURE
DOING WITHOUT KNOWING IS A…

DATA ORIENTED
MACHINE LEARNING
WORKFLOW
3RD APPROACH,FINAL

A COMPUTER PROGRAM
IS SAID TO LEARN FROM EXPERIENCE E
WITH RESPECTTO SOME CLASS OF TASKS T
AND PERFORMANCE MEASURE P
IF ITS PERFORMANCE ATTASKS IN T,
AS MEASURED BY P,
IMPROVES WITH EXPERIENCE E

DATA ML
TASK
PREPARED,INPUT FOR
RESULTS
WITH PERFORMANCE
EXPERIENCE FEEDBACK LOOP
LEARNING,VALIDATING

ML
TASK
CLASSIFICATION
REGRESSION
CLUSTERING
DIMENSIONALITY REDUCTION
ASSOCIATION RULES

ML
METHOD
SUPERVISED LEARNING
UNSUPERVISED LEARNING
REINFORCEMENT LEARNING

+--------------------------------------------------+--------+
| pull_request | price |
+--------------------------------------------------+--------+
| https://github.com/octocat/Hello-World/pull/1347 | 100.00 |
+--------------------------------------------------+--------+
| https://github.com/octocat/Hello-World/pull/1347 | 150.00 |
+--------------------------------------------------+--------+

WHAT IS THE PRICE OF PULL REQUEST
WE NEED TO REVIEW?

+--------------+-------+-----------+--------+
| pull_request | files | all_lines | price |
+--------------+-------+-----------+--------+
| ... | 10 | 1000 | 100.00 |
+--------------+-------+-----------+--------+
| ... | 15 | 2000 | 150.00 |
+--------------+-------+-----------+--------+

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.collections import LineCollection
from sklearn.linear_model import LinearRegression
from sklearn.isotonic import IsotonicRegression
from sklearn.utils import check_random_state
n = 100
x = np.arange(n)
rs = check_random_state(0)
y = rs.randint(-50, 50, size=(n,)) + 50. * np.log(1 + np.arange(n))
ir = IsotonicRegression()
y_ = ir.fit_transform(x, y)
lr = LinearRegression()
lr.fit(x[:, np.newaxis], y)

RECIPE FOR A FAILURE
DON’TYOU KNOWYOUR DATA?

+--------------+-------+-----------+--------+
| pull_request | files | all_lines | price |
+--------------+-------+-----------+--------+
| ... | 10 | 1000 | 100.00 |
+--------------+-------+-----------+--------+
| ... | 15 | 2000 | 150.00 |
+--------------+-------+-----------+--------+
| ... | 15 | 2000 | 50.00 |
+--------------+-------+-----------+--------+

+-----+-------+-----------+------------+--------+
| pr | files | all_lines | diff_lines | price |
+-----+-------+-----------+------------+--------+
| ... | 10 | 1000 | 500 | 100.00 |
+-----+-------+-----------+------------+--------+
| ... | 15 | 2000 | 700 | 150.00 |
+-----+-------+-----------+------------+--------+
| ... | 15 | 2000 | 150 | 50.00 |
+-----+-------+-----------+------------+--------+

FEATURES
CLASSES
INTERFACES
INHERITANCE LEVEL
METHODS CALLS
FUNCTIONS CALLS
AFFERENT COUPLING
EFFERENT COUPLING
LLOC METRIC
LCOM METRIC
…

HOWTO CLUSTER OUR DATASET
AUTOMATED WAY IN THE CONTEXT OF…?

IRIS DATASET
1936,RONALD FISHER

Ronald Fisher

1936 paper

The use of multiple measurements in
taxonomic problems

import numpy as np
import matplotlib.pyplot as plt
from sklearn.cluster import KMeans
from sklearn.datasets import make_blobs
n_samples = 1500
random_state = 170
X, y = make_blobs(n_samples=n_samples, random_state=random_state)
X_varied, y_varied = make_blobs(n_samples=n_samples,
cluster_std=[1.0, 2.5, 0.5],
random_state=random_state)
y_pred = KMeans(n_clusters=3, random_state=random_state).fit_predict(X_varied)

+-----+--------+-------+-----------+------------+--------+
| pr | lang | files | all_lines | diff_lines | price |
+-----+--------+-------+-----------+------------+--------+
| ... | java | 10 | 1000 | 500 | 100.00 |
+-----+--------+-------+-----------+------------+--------+
| ... | java | 15 | 2000 | 700 | 150.00 |
+-----+--------+-------+-----------+------------+--------+
| ... | python | 15 | 2000 | 150 | 50.00 |
+-----+--------+-------+-----------+------------+--------+

FAST ARTIFICIAL
NEURAL NETWORK
CLASSIFICATION

HOWTO CLASSIFY OUR DATASET
AUTOMATED WAY TO FIND JUNIOR/SENIOR
DEVELOPER?

from sklearn.datasets import fetch_mldata
from sklearn.neural_network import MLPClassifier
mnist = fetch_mldata("MNIST original")
# rescale the data, use the traditional train/test split
X, y = mnist.data / 255., mnist.target
X_train, X_test = X[:60000], X[60000:]
y_train, y_test = y[:60000], y[60000:]
# mlp = MLPClassifier(hidden_layer_sizes=(100, 100), max_iter=400, alpha=1e-4,
# solver='sgd', verbose=10, tol=1e-4, random_state=1)
mlp = MLPClassifier(hidden_layer_sizes=(50,), max_iter=10, alpha=1e-4,
solver='sgd', verbose=10, tol=1e-4, random_state=1,
learning_rate_init=.1)
mlp.fit(X_train, y_train)

IN 99.99% APPS
YOU WILL NOT WRITE
ALGO FROM SCRATCH

http://scikit-learn.org/stable/_static/ml_map.png

ML IS NOT
A SINGLE RUN
OF ALGORITHM

ML
PROCESS
DEFINE A PROBLEM
GATHER YOUR DATA
UNDERSTAND YOUR DATA
PREPARE DATA FOR ML
SELECT & RUN ALGO(S)
TUNE ALGO(S) PARAMETERS
SELECT FINAL MODEL
VALIDATE FINAL MODEL

ML
PROCESS
DEFINE A PROBLEM
ANALYZE YOUR DATA
UNDERSTAND YOUR DATA
PREPARE DATA FOR ML
SELECT & RUN ALGO(S)
TUNE ALGO(S) PARAMETERS
SELECT FINAL MODEL
VALIDATE FINAL MODEL

+-----+--------+-------+-----------+------------+--------+----------+
| pr | lang | files | all_lines | diff_lines | price | currency |
+-----+--------+-------+-----------+------------+--------+----------+
| ... | java | 10 | 1000 | 500 | 100.00 | USD |
+-----+--------+-------+-----------+------------+--------+----------+
| ... | java | 15 | 2000 | 700 | 150.00 | USD |
+-----+--------+-------+-----------+------------+--------+----------+
| ... | python | 15 | 2000 | 150 | 50.00 | CAD |
+-----+--------+-------+-----------+------------+--------+----------+

THANKS
mariuszgil
HAPPY LEARNING YOUR MACHINES!

Mariusz Gil "Machine Learning"

More Related Content

Similar to Mariusz Gil "Machine Learning" (20)

More from Fwdays (20)

Recently uploaded (20)

Mariusz Gil "Machine Learning"