20/1/2020 DS4A

LOG OUT

Curriculum Forum

Welcome to DS4A Colombia

Welcome to the 2019 Colombia Data Science for All (DS4A) program. We are thrilled
to welcome you to be part of this unique opportunity to learn data science for the
digital future. Below you will find some useful links for getting setup prior to the start
of the program as well as course materials that will get updated regularly.

Student Welcome Packet

Environment Setup Instructions

Curriculum Outline

Projects Overview

Daily Journal Template

Program Introduction Deck

Learning Resources

––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

PROJECTS

Datathon Project ▼

Datathon Project Materials

Project Work

INTRODUCTORY DECK

PROBLEM STATEMENT

DATASETS

Final Project ▼

Final Project Material: You should also download the Projects Overview
document from the section above.

Project Work

https://ds4a-colombia.correlation1.com/lesson-plan 1/22
20/1/2020 DS4A

INTRODUCTORY DECK LOG OUT

RESOURCES
Curriculum Forum

WEEK 1

Case 20.1: Data Science, Machine Learning, and Artificial Intelligence at a ▼

Glance

What is data science? Machine learning? Artificial intelligence? These terms

are all used interchangeably in our society, as if they are the same thing – but
in reality they are quite different. Before we can answer those questions
though, we have to tackle a much more fundamental one: What is data?
Surprisingly, it is quite hard to come up with a simple definition for it.

Introduction

READING

INTRODUCTION

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 1.1: Identifying Expansion Opportunities for Luxury Commercial Airline ▼

Flights

You are an employee for GrowthAir, a growing commercial airline company. In

the past few years, GrowthAir has expanded luxury flight services to locations
across the globe. Following your team's excellent performance in identifying
new business opportunities last year, you have been tasked with identifying
the top countries to further expand GrowthAir's luxury flight service.

Using git Python Basics

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

https://ds4a-colombia.correlation1.com/lesson-plan 2/22
20/1/2020 DS4A

Case 1.2: How are trading volume and volatility related for energy stocks? ▼
LOG OUT

Curriculum
You are an analyst at a large bank Forum
focused on natural resource stock
investments. This case begins with a brief overview of SOME data, after which
you will: (1) learn how to use the Python library pandas to load the data; (2)
use pandas transform this data into a form amenable for analysis; and finally
(3) use pandas to analyze the above question and come to a conclusion. As
you may have guessed, pandas is an enormously useful library for data
analysis and manipulation.

Data Extraction Data Transformation

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 1.3: Automating the extraction of Financial Data for Stocks ▼

You have recently joined the data science division of a multinational bank.
Over the past month you've been working with a variety of stock data and are
looking to gather fundamental data on a select group of energy stocks. The
firm has no experience doing this in an automated fashion, instead relying on
time-consuming manual labor up to this point.

Data Extraction Web Scraping

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 2.1: Should we develop a commercial SNAP test for predicting recovery ▼
from spinal cord injuries?

You are a consultant for a pharmaceutical company. They would like you to
answer the following: "How well do SNAP tests predict six-month recovery
rates and should they be commercially developed?"

Interpretation of charts, graphs, & tables

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO
https://ds4a-colombia.correlation1.com/lesson-plan 3/22
20/1/2020 DS4A

LOG OUT

Curriculum Forum
Case 2.2: How do users engage with a mobile app for automobiles? ▼

You are a data scientist for a large luxury automobile company. Your company
wants you to uncover behavioral patterns of the users who engage with a
mobile app. They believe that if you can find discernible patterns, your
company can leverage those insights to give users incentives to use the app
more frequently.

Interpretation of charts, graphs, & tables

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 1.4: How are returns and volatility related for energy stocks and the ▼
broader market?

You recently conducted an analysis of several energy stocks and how their
trading volume is related to their volatility. Your boss was quite pleased with
your previous analysis, and now wants you to conduct additional analysis so
he can figure out how to size potential positions in these stocks... i.e. what
percentage of the investment portfolio should be dedicated to each of these
stocks. Specifically, he wants you to look at daily returns and volatility for
each stock as well as for the broader market (i.e. not just the energy sector).

Data Transformation

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 1.5: What patterns exist between energy consumption and generation? ▼

Energy supply and demand is a hotly debated topic across world governments
and political parties. You are an analyst for a new nuclear power plant firm,
and are responsible for discerning patterns in electric power generation and
consumption across different energy sources as well as across sectors of the
U.S. economy in order to help drive business strategy.

Data Visualizations with Python

https://ds4a-colombia.correlation1.com/lesson-plan 4/22
20/1/2020 DS4A

DOWNLOAD CASE LOG OUT

CASE ANSWERS Curriculum Forum

CLASS VIDEO

Week 1 Extended Case: How can we control the increasing number of ▼

accidents in New York?

The city of New York has seen a rise in the number of accidents on the roads
in the city. They would like to know if the number of accidents have increased
in the last few weeks. Your task is to format the given data and provide
visualizations that would answer the specific questions the client has.

Python basics Web Scraping Data Transformation Data Visualizations with Python

DOWNLOAD CASE

CASE ANSWERS

Submissions for Week 1 Extended Case are now closed.

WEEK 2

FICO Case: Is it sound practice to use FICO credit scores to evaluate credit ▼
risk?

Lenders, such as banks and credit card companies, use credit scores to
evaluate the potential risk posed by lending money to consumers and to
mitigate losses due to bad debt. You are an analyst for a large loan agency,
and your role is to help your company make decisions on whether or not to
approve a loan.

Additional Case

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 3.1: What factors are critical for mortgage applications? ▼

https://ds4a-colombia.correlation1.com/lesson-plan 5/22
20/1/2020 DS4A

What information do I have? What information do I need? Do I have enough LOG OUT
information? These are crucial questions that all data scientists must ask
themselves before beginning any analysis. In this case, you will be furthering
Curriculum
your skills in gathering information Forum information sufficiency
and assessing

Information Gathering Information sufficiency

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 4.3: What features determine the price of an Airbnb rental? ▼

Airbnb has rolled out a new service to help listers set prices. Airbnb makes a
percentage commission off of the listings, so they are incentivized to help
listers price optimally; that is, at the maximum possible point where they will
still close a deal. You are an Airbnb consultant helping with this new pricing
service.

Defining Objectives & Questions Business Quantification Selecting Data Visualizations

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 3.2: Spatial and Temporal Trends in Crop Production ▼

Understanding crop production and yield patterns over seasons, years, and
different geographies may yield insights that can improve resilience to food
shortages and, ultimately, improve international food security. You are a
researcher for a think tank that is interested in proposing scientific
investigations and policy solutions based on your insights.

Information Gathering Information sufficiency

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

https://ds4a-colombia.correlation1.com/lesson-plan 6/22
20/1/2020 DS4A

Case 4.4: How do we deploy the Chicago police force to efficiently fight ▼
crime? LOG OUT

Curriculum Forum
So, you found a dataset available to the Chicago PD from 2017 with
information on crimes committed throughout the city. In this case, we will
focus on exploratory analysis to construct some preliminary strategies for
police deployment. These strategies can be further consolidated or dismissed
using more rigorous statistical analysis.

Breaking Problems Down Generating Hypotheses Data Visualization

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Week 2 Extended Case: How is employee happiness related to their risk of ▼

leaving?

Companies generally like to retain their employees, since hiring and training
new employees is costly and risky. The rate at which employees are quitting is
called the turnover rate, and the high cost of employee turnover has been
pointed out extensively in management literature. You work in your company’s
new talent analytics department. Your company has thousands of employees
and is interested in reducing the time spent recruiting and hiring new
employees to replace the people that quit.

Information Sufficiency Breaking Problems Down Data Visualization

DOWNLOAD CASE

CASE ANSWERS

Week 2 Extended Case is for practice and not for grading.

WEEK 3

Case 12.1: How can we optimize our sales of financial products? ▼

You are a data analyst at a large financial services firm that sells a diverse
portfolio of products. In order to make these sales, the firm relies on a call
center where sales agents make calls to current as well as prospective
customers. The company would like you to dive into their data to devise
strategies to increase their revenue or reduce their costs. Specifically, they

https://ds4a-colombia.correlation1.com/lesson-plan 7/22
20/1/2020 DS4A

would like to double down on their most reliable customers, and to cut out
sales agents whom are not producing outcomes. LOG OUT

Relational Databases SQL SQLAlchemy

Curriculum Forum

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 12.2: How can we build a company database to handle product sales ▼
end-to-end?

You are a data analyst for the same large financial services firm as in the
previous case. The firm was pleased with your analysis and now the see the
value of having databases that can easily be queried using SQL. It would
therefore like to move its data, which is currently stored as CSV files, on to a
database.

DBMS PostgresSQL Intro to data storage

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

AWS Setup: Frequently record weather data in your EC2 instance ▼

Set up instructions for AWS

Amazon Web Services EC2

DOWNLOAD CASE

CLASS VIDEO

Case 12.3: Analyzing Net Promoter Score (NPS) data with SQL ▼

You are a data scientist at a new but fast-growing startup. The startup
released its first product 12 months ago and has been tracking Net Promoter
Score (NPS) over its growing customer base since the product's launch. The
startup wants you to investigate the data and answering the following

https://ds4a-colombia.correlation1.com/lesson-plan 8/22
20/1/2020 DS4A

question: "Has our NPS improved over time? And has our average NPS
decreased in specific periods over the last 12 months?" LOG OUT

SQL Postgres AWS RDS

Curriculum Forum

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Week 3 Extended Case: How are different customer cohorts affecting our NPS ▼
ratings over time?

You are a data scientist for the same fast-growing startup in the previous case.
You presented your analysis to the executive team of the startup, and now
they want to dig deeper into various customer segments. They want to zoom
in on specific interesting customers, such as those who have left a wide range
of scores over time.

SQL Postgres AWS

DOWNLOAD CASE

CASE ANSWERS

Submissions for Week 3 Extended Case are now closed.

WEEK 4

Week 4 Environment Setup: Before Class ▼

Please do the following before Friday, 1st November 2019

AWS S3 Dash

ENVIRONMENT

PRECASE

Case 5.1: How do we prepare data for use with an analytics platform? ▼

https://ds4a-colombia.correlation1.com/lesson-plan 9/22
20/1/2020 DS4A

The company has hundreds of thousands of users and has been collecting
data about trips taken on each of their bikes. Since the dataset collected is LOG OUT
quite large and increasing by the day, they have subscribed to a new analytics
platform which gives them information and insights when they feed their trips
Curriculum
data into it. However, the analytics platformForum
requires the collected data to be
cleaned and converted into a certain format, for which the client requires your
help.

Data Wrangling Data Cleaning AWS Storage S3 EC2 data processing

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 4.1: How does my company's sales data compare across different states ▼
over time?

Your company has stores all over the United States. The company has
collected data consisting of line-level order information from all its stores. The
company wishes to compare sales data across different months and
geographies and make this information available to executive members and
key shareholders. If each person has access to an interactive dashboard, it can
offer a foundation for further dialogue and great decision making.

Dash Dashboards Plotly Data Visualization

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 5.2: How do we prepare Yelp data to answer business questions? ▼

Yelp is a very popular website where anyone can write a review about
restaurants, hotels, spas or any place that runs a business. They have decided
to start analysing the data they have to add new features to their website
based on the analysis.

Data Cleaning Data Transformation

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

https://ds4a-colombia.correlation1.com/lesson-plan 10/22
20/1/2020 DS4A

LOG OUT

Case 4.2: How are different types of crimes distributed by district across ▼
time in Boston? Curriculum Forum

You are a data consultant for the Boston Police Department. The department
is looking to optimize its police deployment strategy so that it can tackle the
most crimes as they occur with the fewest number of resources.

Dash Dashboards Plotly AWS Data Visualization

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Week 4 Extended Case: Analyzing an Automated Bitcoin Trading Strategy ▼

As part of a crypto trading desk, you've been tasked with analyzing an

automated Bitcoin trading strategy. The creator of the strategy has made
historical data of the strategy available to you. Your task is to make an
interactive dashboard that your firm can use to visualize the strategy's trades
and performance.

Dash Dashboards Plotly AWS Data Visualization

DOWNLOAD CASE

CASE ANSWERS

Submissions for Week 4 Extended Case are now closed.

WEEK 5

Week 5 Environment Setup: Do this before class ▼

Please do the following before class on Friday, 8th November 2019

Environment Setup

ENVIRONMENT

https://ds4a-colombia.correlation1.com/lesson-plan 11/22
20/1/2020 DS4A

Case 7.1: Do there exist significant differences between the balances of my ▼ OUT
LOG
various customers' cohorts?
Curriculum Forum

You are leading a business analytics unit in a bank and have been asked to
support the marketing unit to conduct a customer segmentation analysis. You
are provided with a dataset comprising a sample of customers, their bank
account balances, and some demographic information about them.

Hypothesis Testing Experiments p-values Statistical Inference

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 7.2: Do the types of crimes committed in Chicago depend on location ▼

and time?

Previously, you investigated crime data for the Chicago police department,
and discovered many potential factors that could be associated with crime
incidents. Now, the police department wants you to finalize your report to
them so that they can start implementing some strategies based on your
findings. However, because deploying a new strategy is resource intensive,
they want you to confirm that the patterns you observed are not merely due to
randomness.

Hypothesis Testing Chi-square test Statistical Inference

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 7.3: How can I experiment with online travel deals to attract more ▼
customers?

Your company TravelSmart is a growing company and is competing with a few

big players in the market. As part of a marketing promotion to increase the
user base, you have decided to give an extra 5% discount for users who book
both flights and hotels at TravelSmart.com, rather than flight tickets alone.
You are interested in learning how this marketing deal affects a variety of
metrics.

A/B Testing

https://ds4a-colombia.correlation1.com/lesson-plan 12/22
20/1/2020 DS4A

DOWNLOAD CASE
LOG OUT
CASE ANSWERS
Curriculum Forum
CLASS VIDEO

Case 7.5: Do prison reform programs reduce recidivism? ▼

Recidivism is a measure of a former prisoner's likelihood to be re-arrested, re-

convicted, or returned to prison with or without a new sentence during a
three-year period following the prisoner's release. Will there be a reduction in
recidivism if the proposed program were to be widely adopted?

Statistical Inference Causal Inference

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 5.3: How do I pre-process text data from Yelp reviews so I can analyze ▼
it?

You are a business consultant for small and medium-sized businesses with a
large number of customers. You would like to help your small businesses
understand what factors are driving positive and negative customer
experiences.

Data Transformation Text Wrangling Natural Language Processing

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Week 5 Extended Case: What are the most important factors driving negative ▼
reviews?

We will investigate a large body of reviews related to movies and television

left on Amazon between 1996 and 2014.

Data Transformation Text Wrangling Natural Language Processing

https://ds4a-colombia.correlation1.com/lesson-plan 13/22
20/1/2020 DS4A

DOWNLOAD CASE LOG OUT

CASE ANSWERS
Curriculum Forum

Submissions for Week 5 Extended Case are now closed.

WEEK 6

Week 6 Environment Setup: Do this before class ▼

Please do the following before class on Friday, 15th November 2019

Environment Setup

ENVIRONMENT

Case 6.1: Is there a pay discrimination between men and women in your ▼
organization?

The human resources department has provided you with an employee

database that contains information on various attributes such as performance,
education, income, seniority, etc. We will use linear regression techniques on
this dataset

Linear Regression Interpreting Linear Models

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 6.2: How should we price homes in Seattle? ▼

You have been hired as a data scientist by a large real estate company in their
Seattle office. Your job is to assist Seattle residents willing to sell their home
with determining an optimal price to sell their property at in order to
maximize their proceeds while still being able to find willing buyers.

Variable Transofrmations Interaction Effects

https://ds4a-colombia.correlation1.com/lesson-plan 14/22
20/1/2020 DS4A

DOWNLOAD CASE
LOG OUT
CASE ANSWERS

CLASS VIDEO Curriculum Forum

Case 6.3: How much should properties be worth in Milwaukee, Wisconsin? ▼

(Part I)

You are a property developer who frequently buys properties. It would be very
useful to get a fair estimate of the price of a property before seeing the
asking price, based on features like its size and location. Your task is to build
a model to predict property prices in the city of Milwaukee, Wisconsin

Residual Analysis Regression

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 6.4: How much should properties be worth in Milwaukee, Wisconsin? ▼

(Part II)

You are the same property developer from the previous case, with the same
goal. Although the previous model you built was a good start, it did not
incorporate all the variables you wished to include, and you are skeptical of
how well it might work on data that it was not trained on.

Colinearity Regression

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Week 6 Extended Case: How can I price an insurance quote across the states ▼
of US?

You are the chief data scientist in a large insurance company and you are
tasked to build an accurate predictive model to understand what factors affect
the claim amount.

https://ds4a-colombia.correlation1.com/lesson-plan 15/22
20/1/2020 DS4A

Regression Modeling

LOG OUT

DOWNLOAD CASE
Curriculum Forum
CASE ANSWERS

Submissions for Week 6 Extended Case are now closed.

WEEK 7

Week 7 Environment Setup: Do this before class ▼

Please do the following before class on Friday, 22nd November 2019

Environment Setup

ENVIRONMENT

Case 9.1: How important is the income source of an online loan applicant? ▼

As a data scientist at an emerging P2P lending company, you must answer the
following question: "Should the company verify the income source of an
online loan applicant before approving their loan?"

Logistic Regression

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 11.1: How can I compare different models that predict the probability ▼
of defaulting on a loan?

Traditional commercial banks typically did not rely on statistical modeling to

decide whether personal loans should be issued, although this is changing
rapidly nowadays. You are a data scientist working in a modern commercial
bank. Your task is to build a default probability model that you feel
comfortable putting into production.

Cross Validation

https://ds4a-colombia.correlation1.com/lesson-plan 16/22
20/1/2020 DS4A

DOWNLOAD CASE LOG OUT

CASE ANSWERS
Curriculum Forum
CLASS VIDEO

Case 9.2: How do I build a prediction model for Lending Club loan defaults? ▼

The biggest question for every P2P lending company IS whether a user will
default or not. Your task is to build a classification model for determining
whether a user will default on their loan or not

Logistic Regression Naive Bayes Classifier Random Forest Cross Validation

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 9.3: Can we detect patterns among specific consumers to identify ▼

potential fraud?

You are working for the fraud team at a large insurance company. Policies
which are issued are kept track of and any filed claims are examined and
evaluated to determine legitimacy and final approval for pay out by the
insurance company. It is the role of the fraud team to determine which filed
claims should be approved and which should be denied.

Clustering Models Unsupervised Learning

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Week 7 Extended Case: Does a job training program improve the earnings of ▼
disadvantaged workers?

In this case we will continue our discussion of causal inference. We will study
the importance of covariate balance and explore how to perform matching to

https://ds4a-colombia.correlation1.com/lesson-plan 17/22
20/1/2020 DS4A

get this balance. We will leverage a lot of the new classification models we
have learned for this purpose. LOG OUT

Causal Inference
Curriculum Forum

DOWNLOAD CASE

CASE ANSWERS

Week 7 Extended case submissions are now closed.

WEEK 8

Week 8 Environment Setup: Do this before class ▼

Please do the following before class on Friday, 29th November 2019

Environment Setup

ENVIRONMENT

Case 10.1: Which salient variables influence the price of homes? ▼

The modern real estate business is increasingly dominated by large quantities

of data. Real estate investment requires sound judgment and accurate price
prediction to determine if a property is a suitable investment. Your task is to
answer the following question: "Which variables are most important for
predicting home prices?"

L1/L2 Regularization

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 17.2: How can we predict the sentiment associated with a customer ▼
interaction?

You have tens of thousands of customers writing reviews on products each

day. Each review contains textual feedback along with a 1-to-5 star rating

https://ds4a-colombia.correlation1.com/lesson-plan 18/22
20/1/2020 DS4A

system (1 being least satisfied and 5 being most satisfied). You also have a
customer support team which interacts with customers over call and LOG OUT
messaging services. our task is to build models which can identify the
sentiment (positive or negative) of each of these non-rated interactions
Curriculum Forum
Sentiment Analysis Natural Language Processing

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 8.1: How do I select survey participants from a pool of potential

▼
customers at my soon-to-be-opened sandwich shop in order to plan the
menu?

Your friends want to open a sandwich shop in an area surrounded by office

buildings. A key aspect of this planning is to get a good grip on what your
potential customers would like to order and what the popular items on the
menu will be. Your task is to devise a plan for how to select people from the
potential customer base to send a survey to, while still obtaining answers
helpful for planning the menu.

Sampling

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Case 18.1: How do I predict the fair market prices of used cars? ▼

There are many companies selling used (refurbished) cars across the United
States. As automobiles depreciate in value as they age, this is an extremely
competitive industry, and we have to price the car right in order to win
business. You are a data scientist tasked with building a predictive model for
the prices of used car sales around the country.

Deep Learning Neural Networks

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

https://ds4a-colombia.correlation1.com/lesson-plan 19/22
20/1/2020 DS4A

LOG OUT
Case 18.2: How does computer vision work? ▼

Curriculum Forum

Image data is everywhere. Everyone is carrying high-resolution cameras

everywhere they go, taking pictures and video and posting them to the
internet. Many social networks manage pictures of their users, and want to
identify said users in order to classify pictures and better understand the
social graph.

Deep Learning Convolutional Neural Network CNN

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

WEEK 9

Week 9 Environment Setup: Use Week 8 Environment ▼

Please use week 8 environment for all classes this week

Environment Setup

WEEK 8 ENV

Case 17.3: Spanish Tweet Classifier ▼

Many social networks analyze the communications that run through them, and
would be interested in classifying them according to whether they exhibit
positive or negative sentiment.

Recurrent Neural Network (RNN) Sentiment Analaysis Natural Language Processing

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

https://ds4a-colombia.correlation1.com/lesson-plan 20/22
20/1/2020 DS4A

Case 18.3: How do I optimize promotions placement on an e-commerce ▼

website? LOG OUT

Curriculum Forum
You want to launch promotions on your website. However, targeting users
effecively is a complex task, and you would like to build a system that can
handle this for you.

Reinforcement Learning

DOWNLOAD CASE

CASE ANSWERS

CLASS VIDEO

Presentation Tips and Guidelines ▼

Some tips and guidelines for your presentations

Guide

DOWNLOAD NOW

WEEK 10

Final Project Submission - Deadline: 10th December 2019, 11:59 PM ▼

Please use this google form to submit your final project. Please add the report
and your project files to a zip folder and submit. Only one submission per
team should be made. Deadline: 10th December 2019, 11:59 PM

Final Project

SUBMIT

Final Project Presentation - Deadline: 10th December 2019, 11:59 PM ▼

https://ds4a-colombia.correlation1.com/lesson-plan 21/22
20/1/2020 DS4A

LOG OUT
Please use this google form to submit your final project presentation. Only
one submission per team should be made. Deadline: 10th December 2019,
11:59PM Curriculum Forum

Final Project Presentation

SUBMIT

Datathon Project Submission - Deadline: 12th December 2019, 11:59 PM ▼

Please use this google form to submit your datathon project report. Please
add the report and your project files to a zip folder and submit. Only one
submission per team should be made.

Datathon

SUBMIT

https://ds4a-colombia.correlation1.com/lesson-plan 22/22