Frequently Asked Questions

The flipped classroom is a pedagogical approach in which the roles that class time and homework typically play in learning are reversed.

Instead of coming to class to listen to a lecture together and doing “homework” individually on your own time, this approach “flips” the model. You listen to lectures and acquire theoretical knowledge individually on your own time. Then you do work and solve problems together in the virtual classroom.

Watch this 2-minute YouTube video to learn more: https://www.youtube.com/watch?v=iQWvc6qhTds

A learning management system, or LMS, is a system used to deploy and track online training. Data Society’s LMS is built using the LearnDash plugin for WordPress. It’s how we get the course materials to you and track your progress. You may also see it referred to as our online learning site, e-learning site, or online learning portal.

Visit https://learn.datasociety.com/recover-password/ and follow the instructions to reset your password.

No. Students may not share their credentials with any other individuals.

The inability to update the profile picture or avatar within a user’s profile is a known issue. There is no solution at this time.

Each lesson in this course includes a number of topics. Some topics are theoretical, while others are more practical. The theoretical topics are followed by quizzes so that you can check your understanding. The practical, coding-based topics are labeled with either (CW) for coding walkthrough or (E) for exercise, as described below:

• (CW). Coding walkthroughs are instructor-led coding walkthroughs of previously discussed topics. You will be provided with the dataset used by the instructor as well as a breakdown of the steps followed.
   
• (E). Exercises are your opportunity to practice coding independently. You will be asked to complete a series of steps in Python using a provided dataset. The results can be checked against a provided answer file.

We recommend that you start with Lesson 1, Topic 1 and progress through the course in a linear fashion, as the topics build on one another.

However, we have not built in restrictions that require you to do so.

It depends. There are 3 hours, 29 minutes, 17 seconds of recorded video within this course’s three self-directed lessons, in addition to outside resources for you to explore, several knowledge checks to confirm your learning, and independent coding exercises. How long the lessons take you to complete will depend on factors such as your baseline knowledge of neural networks and how quickly you code.

Quizzes are short knowledge checks to make sure that you have understood the major concepts in a given topic.

If you get a quiz question wrong, you will be invited to retake the quiz. You may retake the quiz as many times as you like.

Much of the content for this course, including all of the videos, is fully optimized for mobile viewing.

However, topics identified as coding walkthroughs (CW) and exercises (E) require the use of a Python environment. We strongly encourage that you work through these topics using a desktop or laptop.

The course will remain accessible until at least June 1, 2021.

The course forum will remain open until at least June 1, 2021.

However, Data Society’s experts will actively monitor the forums and respond to questions only until the date of the last instructor-led session on December 8, 2020.

Data Society will make its best effort to resolve questions about accessing the LMS and navigating course content as quickly as possible, within regular business hours.

Questions about course content, including troubleshooting in Python, should be answered within a 48-hour period.

If you see a question on the forum that you know the answer to, we welcome timely, accurate crowdsourced solutions.

Email training@datasociety.com for additional support.

We have designed and sequenced this course to give you the best possible chance of success in meeting your learning objectives. We recommend that you complete every topic and activity prior to the instructor-led session, as substantial lecture-style review of the material is not planned.

The instructor-led session has two objectives:

(1) to answer your questions, and

(2) to give you the opportunity to apply the concepts you have learned to your own dataset with the benefit of an instructor’s and classmates’ insights.

Therefore, there’s a few things that you can do to prepare:

• Make note of any outstanding questions that you have
• Identify a dataset on which you’d like to practice and have it with you.
• Have your computer accessible, with Python installed for in-class coding work. 
• Download the Zoom client, if you are able. We recommend using the Zoom client to maximize the tool’s functionality, but understand if students must use the browser-based version. You can read more about Zoom and its functionality here.

You are free to choose any dataset that is of interest to you. However, there are a few ideal characteristics:

• Is publicly available
• Can be shared with your fellow classmates
• Of a relevant subject matter (e.g., engineering, materials science, etc.)
• Contains > 1,000 samples
• Includes numeric data
• Is multivariate (more than 5 variables)
• Is well suited for either classification or regression (meaning you are looking for a binary target, or know what element you want to predict)

We have identified three datasets for students who are unable to locate a dataset that meets the requirements. They are:

• Steel Plates Faults: This multivariate dataset contains 1,941 instances of steel plate faults, classified into 7 different types.
• Concrete Compressive Strength: This multivariate dataset contains 1,030 instances and is good for regression analysis. 
• Superconductivity Data: This multivariate dataset contains 21,263 instances of data on superconductors and their relevant features.

Use this link to download all three.

You may also simulate a dataset, if this is a technique familiar to you. You can read more on the topic here.

The session will be conducted using the Zoom platform.  You can read more about Zoom and its functionality here.

You should have received a calendar invite with a link to the Zoom meeting on Monday, November 30. If you did not receive this information, email training@datasociety.com.

Anaconda and package install instructions are provided here. The required pip_env.txt file is located here.

The course materials (all videos, for example) are all done using Jupyter Notebook and the installation instructions are made to support that.

However, for the purposes of this course the code works for either without any functional differences. You can therefore choose to use either use Notebook or Lab for the course.

absl-py==0.9.0
astunparse==1.6.3
attrs==20.2.0
cachetools==4.1.1
certifi==2020.6.20
chardet==3.0.4
colorama==0.4.4
cycler==0.10.0
dill==0.3.2
dm-tree==0.1.5
future==0.18.2
gast==0.3.3
google-auth==1.22.1
google-auth-oauthlib==0.4.2
google-pasta==0.2.0
googleapis-common-protos==1.52.0
grpcio==1.33.1
h5py==2.10.0
idna==2.10
importlib-resources==3.2.1
joblib==0.17.0
Keras-Preprocessing==1.1.2
keras-tuner==1.0.1
kiwisolver==1.3.0
Markdown==3.3.3
matplotlib==3.3.2
numpy==1.18.5
oauthlib==3.1.0
opt-einsum==3.3.0
pandas==1.1.3
Pillow==8.0.1
promise==2.3
protobuf==3.13.0
pyasn1==0.4.8
pyasn1-modules==0.2.8
pyparsing==2.4.7
python-dateutil==2.8.1
pytz==2020.1
requests==2.24.0
requests-oauthlib==1.3.0
rsa==4.6
scikit-learn==0.23.2
scipy==1.5.3
seaborn==0.11.0
six==1.15.0
tabulate==0.8.7
tensorboard==2.3.0
tensorboard-plugin-wit==1.7.0
tensorflow==2.3.1
tensorflow-datasets==4.0.1
tensorflow-estimator==2.3.0
tensorflow-metadata==0.24.0
termcolor==1.1.0
terminaltables==3.1.0
threadpoolctl==2.1.0
tqdm==4.51.0
urllib3==1.25.11
Werkzeug==1.0.1
wincertstore==0.2
wrapt==1.12.1

• Issue: “conda activate nasa-nn” only works in conda 4.8 and later. Solution: “conda update conda” will provide the needed update.
• Issue: Participants do not have a “nasa-nn” folder on their Desktop when they reach the “cd Desktop/nasa-nn” step. Solution: Create one! This is where you will download the materials for each lesson.
• Issue: tensorflow wasn’t installed by the pip_env.txt list. Solution: see “What issues have been reported while trying to complete the exercises?” below on this page.

• Issue: In the Lesson 1, Topic 5 notebook, tensorflow wasn’t installed by the pip_env.txt list.
  • Solution: The script calls for kiwisolver=1.3.0, and the install was trying to build it from source and failing, so subsequent packages weren’t installed.
    1. Delete the nasa-nn environment.
    2. Perform a conda update on conda, anaconda, python and –all on the base environment.
    3. Recreate the nasa-nn environment (see slide 8).
    4. Change pip_env.txt to install kiwisolver=1.3.1.
    5. Re-run the pip environment load (see slide 10).

If you have any questions trying to install software or packages, contact Herb Schilling (hschilling@nasa.gov).