1. Week 3 (June 6 – June10 , 2016) Summary :
Project 7: Modeling Social Network Structures and their Dynamic Evolution with User- generated Data on IoT. REU Student: Morzouk Lawal Graduate mentors: Marzieh Edraki Faculty Mentor(s): Dr. GuoJun Qi
Week 3 (June 6 – June10 , 2016)
Summary :
Aim:
Distill useful information from user-generated content like semantic segmentation from user generated images and videos.
The model will then be pre-trained on ImageNet: A benchmark dataset with billions of images.
Concepts required:
Machine Learning, LSTM(Long Short Term Memory), Deep Learning and Theano.
Knowledge in C++, Python, and MATLAB.

2. Project 7: Modeling Social Network Structures and their Dynamic Evolution with User-generated Data on IoT. REU Student: Morzouk Lawal Graduate mentors: Marzieh Edraki Faculty Mentor(s): Dr. GuoJun Qi
Accomplishments:
Read 2 research papers on Image Classification.
Figured out how to change the datasets in the code I was given so the program can be trained using ImageNet.
Downloaded some synsets of Images from ImageNet.

3. Summary of Image Classification Paper:
Project 7: Modeling Social Network Structures and their Dynamic Evolution with User-generated Data on IoT. REU Student: Morzouk Lawal Graduate mentors: Marzieh Edraki Faculty Mentor(s): Dr. GuoJun Qi
Summary of Image Classification Paper:
The task in Image Classification is to basically take an array of pixels that represent a single image and assign a label to it. A complete Pipeline can be formalized as follows:
Input: The input consist of a set of N images, each labeled with one of K different classes. This is considered a training set.
Learning: The task is to use the training set to learn what every one task looks like. This step is called Learning a model or Training a classifier. 
Evaluation: In the end, we evaluate the quality of the classifier by asking it to predict labels for a new set of images that it has never been before. We will then compare the true labels of these images to the ones predicted by the Classifier. Intuitively, we’re hoping the predictions match with the true answers (this is called the Ground Truth)

4. Issues/problems encountered and solutions (if any):
Project 7: Modeling Social Network Structures and their Dynamic Evolution with User-generated Data on IoT. REU Student: Morzouk Lawal Graduate mentors: Marzieh Edraki Faculty Mentor(s): Dr. GuoJun Qi
Issues/problems encountered and solutions (if any):
At first, I was unable to find the synsets of images from ImageNet
After finding the synsets, I was unable to download them because I did not have permission to do so.
After obtaining permission, I was able to download a few which only included the validation set.

5. Project 7: Modeling Social Network Structures and their Dynamic Evolution with User-generated Data on IoT. REU Student: Morzouk Lawal Graduate mentors: Marzieh Edraki Faculty Mentor(s): Dr. GuoJun Qi
Plans for next week:
Hopefully download all images for Training set.
Implement the images in my code and start training the code with them.