Data Source

```python import numpy as np import pandas as pd import matplotlib.pyplot as plt plt.style.use(‘seaborn’)

Download and extract data from: https://data.worldbank.org/country/philippines (Both CSV and EXCEL for demo)

Necessary Imports

Walkthrough of datacamp tutorial of : (https://www.datacamp.com/community/tutorials/deep-learning-python)

```python import tensorflow as tf import numpy as np import matplotlib.pyplot as plt

Updating Tensorflow and Building Keras from Github

“SUPPOSING that Truth is a woman–what then? Is there not ground for suspecting that all philosophers, in so far as they have been dogmatists, have failed to understand women–that the terrible seriousness and clumsy importunity with which they have usually paid their addresses to Truth, have been unskilled and unseemly methods for winning a woman?” - Nietzsche

Step 1: Organize Imports

A Keras Tutorial on Image Classification With Little Data

A Keras Tutorial on Image Classification With Little Data

Face recognition with Xception

(Inspired by: https://gogul09.github.io/software/flower-recognition-deep-learning)

Face recognition with VGG16

(Inspired by: https://gogul09.github.io/software/flower-recognition-deep-learning)

Face recognition with VGG19

(Inspired by: https://gogul09.github.io/software/flower-recognition-deep-learning)

Face recognition with ResNet50

(Inspired by: https://gogul09.github.io/software/flower-recognition-deep-learning)

Face recognition with Inceptionv3

(Inspired by: https://gogul09.github.io/software/flower-recognition-deep-learning)

Step 1: Install FFMPEG

Step 1: Organize imports

Step 1: Organize imports

Step 1: Organize imports

Using Keras’ Data Augmentation

Using Keras’ Data Augmentation

Step 1: Organize imports

Step 1: Organize Imports

Step 1: Organize Imports

Source: http://cs231n.github.io/python-numpy-tutorial/

Step 1: Organize imports

Step 1: Organize imports

Step 1: Organize imports

Deep Learning Hello World! (LeNet-5)

Assignment 1-a: Deep Learning Hello World! (Baseline)

Assignment 1-d: Deep Learning Hello World! (3-layer MLP+Dropout+Optimizer)

Assignment 1-c: Deep Learning Hello World! (3-layer MLP+Dropout)

Assignment 1-b: Deep Learning Hello World! (3-layer MLP)

Deep Learning (from Tensorflow Tutorial)

Deep Learning (in Tensorflow)

Tensors are multilinear maps from vector spaces to the real numbers.

Running the ‘Hello World!’ routine to check environment settings:

• Clone OpenCV repository to your machine.

excerpts from opencv3 blueprints

excerpts from opencv3 blueprints

Load octave extension

Machine Learning Activity 3

SIGSLAB I

IPython magic mode

Here is a simple and portable deep learning C library libdeep by Bob Mottram.

Caffe image dataset requires train.txt and val.txt which are text listing all the files and their labels.

This task can be accomplished by either of the following methods:

Install dependency

Basic Logistic Regression of HDF5 Data using Caffe by Yangqing Jia

This iPython notebook is based on the Caffe classification.ipynb example using the pre-trained CaffeNet model, an ILSVRC12 image classifier.

“Our face verification method reaches an accuracy of 97.35% on the Labeled Faces in the Wild (LFW) dataset, reducing the error of the current state of the art by more than 27%, closely approaching human-level performance.”
Yaniv Taigman et al.

“Torch7 is a versatile numeric computing framework and machine learning librar that extends Lua. Its goal is to provide a flexible environment to design and train learning machines. Flexibility is obtained via Lua, an extremely lightweight scripting language. High performance is obtained via efficient OpenMP/SSE an CUDA implementations of low-level numeric routines. Torch7 can easily be interfaced to third-party software thanks to Lua’s light interface.”
Ronan Collobert et al.

“Coming up with features is difficult, time-consuming, and requires expert knowledge. When working applications of learning, we spend a lot of time tuning the features. However, these features can be learned.”

“In this paper we address three different computer vision tasks using a single basic architecture: depth prediction, surface normal estimation, and semantic labeling. We use a multiscale convolutional network that is able to adapt easily to each task using only small modifications, regressing from the input image to the output map directly. Our method progressively refines predictions using a sequence of scales, and captures many image details without any superpixels or low-level segmentation. We achieve state-of-the-art performance on benchmarks for all three tasks.”
David Eigen and Rob Fergus

“SVMs are wonderful as a generic classification method with beautiful math behind them. But in the end, they are nothing more than simple two-layer systems. The first layer can be seen as a set of units (one per support vector) that measure a kind of similarity between the input vector and each support vector using the kernel function. The second layer linearly combines these similarities.”

“Experience with real-world data, however, soon convinces one that both stationarity and Gaussianity are fairy tales invented for the amusement of undergraduates.”
David Thomson

“Research is to see what everybody else has seen, and to think what nobody else has thought.”
Albert Szent-Gyorgyi

“Moment is a word which most often refers to an ambiguously short length of time, but also signifies in mathematics a quantitative measure of the shape of a set of points, and in physics relates to the perpendicular distance from a point to a line or a surface.”
Wikiquote

“Could you take my picture… Cause I won’t remember.”
Filter

“The mind of the painter must resemble a mirror, which always takes the colour of the object it reflects and is completely occupied by the images of as many objects as are in front of it. Therefore you must know, Oh Painter! that you cannot be a good one if you are not the universal master of representing by your art every kind of form produced by nature. And this you will not know how to do if you do not see them, and retain them in your mind.”
Leonardo Da Vinci

“The fact that objects in the world appear in different ways depending on the scale of observation has important implications if one aims at describing them. It shows that the notion of scale is of utmost importance when processing unknown measurement data by automatic methods.”
Tony Lindeberg

“They who are acquainted with the present state of the theory of Symbolical Algebra, are aware, that the validity of the processes of analysis does not depend upon the interpretation of the symbols which are employed, but solely upon the laws of their combination.”
George Boole

“Our life is frittered away by detail… Simplicity, simplicity, simplicity! I say, let your affairs be as two or three, and not a hundred or a thousand; instead of a million count half a dozen, and keep your accounts on your thumb nail. In the midst of this chopping sea of civilized life, such are the clouds and storms and quicksands and thousand-and-one items to be allowed for, that a man has to live, if he would not founder and go to the bottom and not make his port at all, by dead reckoning, and he must be a great calculator indeed who succeeds. Simplify, simplify. Instead of three meals a day, if it be necessary eat but one; instead of a hundred dishes, five; and reduce other things in proportion.”
Henry David Thoreau

“Images are information rich, yet while humans interpret them effortlessly, doing so algorithmically remains, paradoxically, hard.”
Stefan van der Walt, founder of scikit-image

“My greatest concern was what to call it. I thought of calling it ‘information,’ but the word was overly used, so I decided to call it ‘uncertainty.’ When I discussed it with John von Neumann, he had a better idea. Von Neumann told me, ‘You should call it entropy, for two reasons. In the first place your uncertainty function has been used in statistical mechanics under that name, so it already has a name. In the second place, and more important, no one really knows what entropy really is, so in a debate you will always have the advantage.”
Claude Elwood Shannon

“My research areas include computer vision and computer graphics. My particular interests are in using vision to automatically build 3-D models from images, computational photography, and image-based rendering…”
Richard Szeliski

“I am, somehow, less interested in the weight and convolutions of Einstein’s brain than in the near certainty that people of equal talent have lived and died in cotton fields and sweatshops.”
Stephen Jay Gould

“Numerical analysis has always been the black sheep of mathematics. I like to say that mathematics is the art of avoiding computation. In pure mathematics if you have to do a computation then it’s because your model is not sophisticated enough, or not elegant enough, or you haven’t got the right abstraction. It was just different, it was a new subject and most of the mathematicians at Stanford, I’d say, were happy to see it get out of the department and go its own way elsewhere.”
Cleve Moler, cofounder of MathWorks

“All my early memories are of forms and shapes and textures. Moving through and over the West Riding landscape with my father in his car, the hills were sculptures; the roads defined the forms. Above all, there was the sensation of moving physically over the contours of foulnesses and concavities, through hollows and over peaks – feeling, touching, seeing, through mind and hand and eye. This sensation has never left me. I, the sculptor, am the landscape. I am the form and I am the hollow, the thrust and the contour.”
Barbara Hepworth

“Many computational applications naturally involve not just a set of items, but also a set of connections between pairs of those items. Is there a way to get from one item to another by following the connections? How many other items can be reached from a given item? What is the best way to get from this item to the other item?.”
Robert Sedgewick

The Zen of Python

“The question isn’t, ‘What do we want to know about people?’, It’s, ‘What do people want to tell about themselves?’”
Mark Zuckerberg

“You could not step twice into the same river.”
Heraclitus of Ephesus

“The problems of image segmentation and grouping remain great challenges for computer vision. Since the time of the Gestalt movement in psychology, it has been known that perceptual grouping plays a powerful role in human visual perception.”
Pedro Felzenszwalb

Input:

“Einstein’s theory of relativity does a fantastic job for explaining big things. Quantum mechanics is fantastic for the other end of the spectrum - for small things. The big problem is that each theory is great for each realm, but when they confront each other, they are ferocious antagonists, and the mathematics falls apart.”
Brian Greene

“I think the next century will be the century of COMPLEXITY.”
Stephen Hawking

“All knowledge can be thought of as either declarative or imperative. Declarative knowledge is composed of statements of fact. Unfortunately, it doesn’t tell us how to find that certain fact. Imperative knowledge is the HOW TO knowledge, or recipes for deducing information.”
John Guttag

“All the phenomena of the universe are presumably in some way continuous; and certain facts, plucked as it were from the very heart of nature, are likely to be of use in our gradual discovery of facts which lie deeper still.”
William Crookes

“The impressionistic method leads into a complete splitting and dissolution of all areas involved in the composition, and color is used to create an overall effect of light. The color is, through such a shading down from the highest light in the deepest shadows, sacrified an degraded to a (black-and-white) function. This leads to the destructions of the color as color.”
Hans Hofmann

“Simplicity is the final achievement. After one has played a vast quantity of notes and more notes, it is simplicity that emerges as the crowning reward of art.”
Frédéric Chopin

“Superpixel segmentation algorithms can be very useful as a preprocessing step for computer vision applications like object class recognition and medical image segmentation. To be useful, such algorithms should output high quality superpixels. There are few superpixel algorithms that can offer this and scale up for practical applications that deal with images greater than 0.5 million pixels. We present a novel O(N) complexity superpixel segmentation algorithm that is simple to implement and outputs better quality superpixels for a very low computational and memory cost…”
Radhakrishna Achanta et al.

“All our knowledge has its origins in our perceptions.”
Leonardo da Vinci