Project Overview Welcome to the Convolutional Neural Networks (CNN) project in the AI Nanodegree! In this project, you will learn how to build a pipeline that can be used within a web or mobile app to process real-world, user-supplied images. Given an image of a dog, your algorithm will identify an estimate of the canine’s breed. If supplied an image of a human, the code will identify the resembling dog breed. Sample Output Along with exploring state-of-the-art CNN models for classification, you will make important design decisions about the user experience for your app. Our goal is that by completing this lab, you understand the challenges involved in piecing together a series of models designed to perform various tasks in a data processing pipeline. Each model has its strengths and weaknesses, and engineering a real-world application often involves solving many problems without a perfect answer. Your imperfect solution will nonetheless create a fun user experience! Project Instructions Instructions Clone the repository and navigate to the downloaded folder. git clone https://github.com/udacity/dog-project.git cd dog-project Download the dog dataset. Unzip the folder and place it in the repo, at location path/to/dog-project/dogImages. Download the human dataset. Unzip the folder and place it in the repo, at location path/to/dog-project/lfw. If you are using a Windows machine, you are encouraged to use 7zip to extract the folder. Download the VGG-16 bottleneck features for the dog dataset. Place it in the repo, at location path/to/dog-project/bottleneck_features. (Optional) If you plan to install TensorFlow with GPU support on your local machine, follow the guide to install the necessary NVIDIA software on your system. If you are using an EC2 GPU instance, you can skip this step. (Optional) If you are running the project on your local machine (and not using AWS), create (and activate) a new environment. Linux (to install with GPU support, change requirements/dog-linux.yml to requirements/dog-linux-gpu.yml): conda env create -f requirements/dog-linux.yml source activate dog-project Mac (to install with GPU support, change requirements/dog-mac.yml to requirements/dog-mac-gpu.yml): conda env create -f requirements/dog-mac.yml source activate dog-project NOTE: Some Mac users may need to install a different version of OpenCV conda install --channel https://conda.anaconda.org/menpo opencv3 Windows (to install with GPU support, change requirements/dog-windows.yml to requirements/dog-windows-gpu.yml): conda env create -f requirements/dog-windows.yml activate dog-project (Optional) If you are running the project on your local machine (and not using AWS) and Step 6 throws errors, try this alternative step to create your environment. Linux or Mac (to install with GPU support, change requirements/requirements.txt to requirements/requirements-gpu.txt): conda create --name dog-project python=3.5 source activate dog-project pip install -r requirements/requirements.txt NOTE: Some Mac users may need to install a different version of OpenCV conda install --channel https://conda.anaconda.org/menpo opencv3 Windows (to install with GPU support, change requirements/requirements.txt to requirements/requirements-gpu.txt): conda create --name dog-project python=3.5 activate dog-project pip install -r requirements/requirements.txt (Optional) If you are using AWS, install Tensorflow. sudo python3 -m pip install -r requirements/requirements-gpu.txt Switch Keras backend to TensorFlow. Linux or Mac: KERAS_BACKEND=tensorflow python -c "from keras import backend" Windows: set KERAS_BACKEND=tensorflow python -c "from keras import backend" (Optional) If you are running the project on your local machine (and not using AWS), create an IPython kernel for the dog-project environment. python -m ipykernel install --user --name dog-project --display-name "dog-project" Open the notebook. jupyter notebook dog_app.ipynb (Optional) If you are running the project on your local machine (and not using AWS), before running code, change the kernel to match the dog-project environment by using the drop-down menu (Kernel > Change kernel > dog-project). Then, follow the instructions in the notebook. NOTE: While some code has already been implemented to get you started, you will need to implement additional functionality to successfully answer all of the questions included in the notebook. Unless requested, do not modify code that has already been included. Evaluation Your project will be reviewed by a Udacity reviewer against the CNN project rubric. Review this rubric thoroughly, and self-evaluate your project before submission. All criteria found in the rubric must meet specifications for you to pass. Project Submission When you are ready to submit your project, collect the following files and compress them into a single archive for upload: The dog_app.ipynb file with fully functional code, all code cells executed and displaying output, and all questions answered. An HTML or PDF export of the project notebook with the name report.html or report.pdf. Any additional images used for the project that were not supplied to you for the project. Please do not include the project data sets in the dogImages/ or lfw/ folders. Likewise, please do not include the bottleneck_features/ folder.

The WashU Physics Machine Learning Club puts jupyter notebooks we will be using internally in this repo. We do not claim authorship of the notebooks.

In this repo I will show you how to embed your Power BI report in Python Jupyter Notebook

This repo contains notebooks from my machine learning sessions. It might not be organized but it will take you from ML basics. Again, they are combination of jupyter notebooks from the classes i teach.

This repository contains all Python scripts and Jupyter Notebooks that were used in the final stages of my undergraduate thesis. To summarise, the thesis involved the analysis of Stage N2 portions of 8 hour EEG recordings across 15 patients in order to extract 'sleep spindles'. The 'Quadratic Parameter Sinusoid' or QPS (Palliyali et. al. 2015) was used as a way to extract 6 quadratic polynomial coefficients that served as statistical descriptors of features of the extracted spindles (and non-spindles) such as their amplitudes, envelope symmetry, frequency, phase and more. The way this was achieved was using non-linear least squares (NLLS) via the Levenberg-Marquadt Algorithm (LM) as a way to perform a best fit of the model to the raw captured spindle. The main goal of the thesis was to use these 6 parameters as learning features in a simple feed-forward neural network in order to classify whether or not an acquired raw portion of an EEG signal is a spindle or not. The conclusion to the study showed that while the QPS model was a great way to reconstruct spindles and extract valuable coefficient data, there is no guarantee the non-linear regression will work since parameter initialisation is highly dependent on whether or not it is known (for certain) if the acquired raw section of the EEG is a spindle or not.

Part I - WeatherPy In this example, you’ll be creating a Python script to visualize the weather of 500+ cities across the world of varying distance from the equator. To accomplish this, you’ll be utilizing a simple Python library, the OpenWeatherMap API, and a little common sense to create a representative model of weather across world cities. Your first objective is to build a series of scatter plots to showcase the following relationships: Temperature (F) vs. Latitude Humidity (%) vs. Latitude Cloudiness (%) vs. Latitude Wind Speed (mph) vs. Latitude After each plot add a sentence or too explaining what the code is and analyzing. Your next objective is to run linear regression on each relationship, only this time separating them into Northern Hemisphere (greater than or equal to 0 degrees latitude) and Southern Hemisphere (less than 0 degrees latitude): Northern Hemisphere - Temperature (F) vs. Latitude Southern Hemisphere - Temperature (F) vs. Latitude Northern Hemisphere - Humidity (%) vs. Latitude Southern Hemisphere - Humidity (%) vs. Latitude Northern Hemisphere - Cloudiness (%) vs. Latitude Southern Hemisphere - Cloudiness (%) vs. Latitude Northern Hemisphere - Wind Speed (mph) vs. Latitude Southern Hemisphere - Wind Speed (mph) vs. Latitude After each pair of plots explain what the linear regression is modelling such as any relationships you notice and any other analysis you may have. Your final notebook must: Randomly select at least 500 unique (non-repeat) cities based on latitude and longitude. Perform a weather check on each of the cities using a series of successive API calls. Include a print log of each city as it’s being processed with the city number and city name. Save a CSV of all retrieved data and a PNG image for each scatter plot. Part II - VacationPy Now let’s use your skills in working with weather data to plan future vacations. Use jupyter-gmaps and the Google Places API for this part of the assignment. Note: if you having trouble displaying the maps try running jupyter nbextension enable --py gmaps in your environment and retry. Create a heat map that displays the humidity for every city from the part I of the homework. heatmap Narrow down the DataFrame to find your ideal weather condition. For example: A max temperature lower than 80 degrees but higher than 70. Wind speed less than 10 mph. Zero cloudiness. Drop any rows that don’t contain all three conditions. You want to be sure the weather is ideal. Note: Feel free to adjust to your specifications but be sure to limit the number of rows returned by your API requests to a reasonable number. Using Google Places API to find the first hotel for each city located within 5000 meters of your coordinates. Plot the hotels on top of the humidity heatmap with each pin containing the Hotel Name, City, and Country. hotel map As final considerations: Create a new GitHub repository for this project called API-Challenge (note the kebab-case). Do not add to an existing repo You must complete your analysis using a Jupyter notebook. You must use the Matplotlib or Pandas plotting libraries. For Part I, you must include a written description of three observable trends based on the data. You must use proper labeling of your plots, including aspects like: Plot Titles (with date of analysis) and Axes Labels. For max intensity in the heat map, try setting it to the highest humidity found in the data set. Hints and Considerations The city data you generate is based on random coordinates as well as different query times; as such, your outputs will not be an exact match to the provided starter notebook. You may want to start this assignment by refreshing yourself on the geographic coordinate system. Next, spend the requisite time necessary to study the OpenWeatherMap API. Based on your initial study, you should be able to answer basic questions about the API: Where do you request the API key? Which Weather API in particular will you need? What URL endpoints does it expect? What JSON structure does it respond with? Before you write a line of code, you should be aiming to have a crystal clear understanding of your intended outcome. A starter code for Citipy has been provided. However, if you’re craving an extra challenge, push yourself to learn how it works: citipy Python library. Before you try to incorporate the library into your analysis, start by creating simple test cases outside your main script to confirm that you are using it correctly. Too often, when introduced to a new library, students get bogged down by the most minor of errors – spending hours investigating their entire code – when, in fact, a simple and focused test would have shown their basic utilization of the library was wrong from the start. Don’t let this be you! Part of our expectation in this challenge is that you will use critical thinking skills to understand how and why we’re recommending the tools we are. What is Citipy for? Why would you use it in conjunction with the OpenWeatherMap API? How would you do so? In building your script, pay attention to the cities you are using in your query pool. Are you getting coverage of the full gamut of latitudes and longitudes? Or are you simply choosing 500 cities concentrated in one region of the world? Even if you were a geographic genius, simply rattling 500 cities based on your human selection would create a biased dataset. Be thinking of how you should counter this. (Hint: Consider the full range of latitudes). Once you have computed the linear regression for one chart, the process will be similar for all others. As a bonus, try to create a function that will create these charts based on different parameters. Remember that each coordinate will trigger a separate call to the Google API. If you’re creating your own criteria to plan your vacation, try to reduce the results in your DataFrame to 10 or fewer cities. Lastly, remember – this is a challenging activity. Push yourself! If you complete this task, then you can safely say that you’ve gained a strong mastery of the core foundations of data analytics and it will only go better from here. Good luck!