The guidance for using the online entry platform and the EOI submission requirements for the Forecasting Prize and the Field Implementation Grant are defined below.
How to Access the Entry Platform:
All EOIs must be submitted through the online entry platform to compete in the Intelligent Forecasting Competition. When you click the link to enter, you will be directed to the login page. If you have an existing account, type in your login information and click LOG IN. You will be redirected to the Intelligent Forecasting entry form.
If you do not yet have an login, click SIGN UP in the upper right corner. When you have registered for an account, you will be redirected to the Intelligent Forecasting entry form.
If at any point you are directed to the Innovation Competitions landing page, find the Intelligent Forecasting Competition and click MORE> to access the entry form.
If at any point you are directed to the Applications page, you can either access a previous entry to the Intelligent Forecasting Competition (if one exists) or you can click Programs in the top bar to return to the Innovation Competitions landing page. Find the Intelligent Forecasting Competition and click MORE> to access the entry form.
For detailed guidance on how to access the Intelligent Forecasting Competition entry intake platform for first time users and existing users, click here. If you have other questions about SMApply, please contact email@example.com.
Forecasting Prize: Submission Requirements
1) The predicted consumption (stock_distributed) values must be submitted in the ‘predicted_value’ field of the submission_format.csv template file. The template is provided in Annex I of the Competition Call (page 27), and can be downloaded using pagthis link. The fields in this file include: year (YYYY); month (MM); site_code; product_code; predicted_value. The predictions should be made monthly for three months: October 2019, November 2019, and December 2019.
2) Any submitted model(s) must be uploaded as a Jupyter Notebook (file extension .ipynb). There are instructions on how to create a Jupyter Notebook in Annex II of the Competition Call (page 28). For USAID learning purposes, please include comments using Jupyter Notebook that clearly document each step taken in the development of the model. These may potentially include, but are not limited to, decisions to include/exclude certain variables, results of tests conducted to assess accuracy, etc.
3) Provide narrative answers to the questions listed below to describe each of the models submitted. While completing the narrative answers is required, answers will not be considered in the evaluation and determination of the prize awards. These answers will be used by USAID to better understand the elements of a successful model.
Competitors will have the opportunity to indicate whether the information included in the narrative response is proprietary or confidential, and provide the information USAID can share publicly or with other competitors. Competitors may include the same or duplicate content.
A. Help USAID to understand the model:
1. Provide a brief, high level summary of the submitted model.
2. Provide the three most important blocks of code from the model. Paste each block below and explain what it does and why it is important to the model.
3. How was the model trained (e.g. hyperparameters, training protocols, specialized hardware)?
4. Please share any relevant charts, graphs, or visuals pertaining to model submission and/or performance. (Optional)
5. Is there anything that was tried but did not make it into the final submission? If so, please explain briefly. (Optional)
6. Is there anything USAID should know regarding model performance, potential issues or quirks, and/or biases (including, but not limited to gender bias) inherent to the proposed model?
B. Recommend future steps:
1. Describe any strengths or limitations to implement this model in the field.
2. If this model is implemented in the field, are there additional recommendations for data, resources or techniques to try?
3. Please share any questions or concerns about the data provided for this competition.
C. Sharing the Model:
1. Can details from the submitted model(s) and narrative (excluding confidential and proprietary information) be included in published research, blogs, or other public forms of learning from this competition?
2. Can the model submitted be shared with other competitors?
1. For competitors that intend to enter the Forecasting Prize but NOT the Field Implementation Grant:
Field Implementation Grant
The goal of the Field Implementation Grant is to work with stakeholders in Côte d’Ivoire to customize, continually improve, and field test the intelligent forecasting model. The grant winner will work with selected facilities in Côte d’Ivoire to use the model to improve the accuracy of contraceptive consumption forecasting and ordering, and the utility of the model for facilities.
During the approximately one year or longer implementation period, the expected activities may include, but are not limited to:
USAID encourages competitors to propose additional activities that are relevant to the successful implementation of an intelligent forecasting model. The successful competitor will also be expected to provide, to stakeholders in Côte d’Ivoire and USAID/Washington, DC: regular updates, relevant data, and a final report on their results, implementation process, and lessons learned to contribute to the body of knowledge around using intelligent methods to forecast consumption of contraceptives.
Field Implementation Grant: Submission Requirements
A. Responses to additional questions on the online entry platform that address the recommended field implementation approach. Responses should be clear, thorough, and address the following questions:
1. Understanding of the Problem: Please articulate the problem of contraceptive forecasting and your overall understanding of why and how an intelligent method might be utilized, particularly in Côte d’Ivoire. (500 words)
2. Local Applicability: Please describe how the proposed model will address local needs and context in the Côte d’Ivoire public sector health system. (500 words)
3. Implementation Capacity: Please indicate your capacity to deliver the project, including details of how you will staff and manage the grant. (500 words)
4. Past Performance: Please describe your past experience implementing similar projects. (250 words)
B. Describe your willingness to collaborate. This section will not be evaluated.
1. Are you willing to share your field implementation approach with other competitors?
2. Are you willing to incorporate another competitor's intelligent forecasting model into your field implementation approach?
3. Are you willing to partner with another competitor through a co-creation process?
Co-Creation and Concept Notes
USAID will test models and evaluate EOIs to develop a shortlist of competitors for the Forecasting Prize and Field Implementation Grant, who will be invited to a virtual co-creation effort. Competitors who only submitted an EOI to the Forecasting Prize must state they are open to partnering with a field implementation team to be eligible to participate in co-creation activities.
During co-creation, shortlisted competitors will have the opportunity to improve their models and field implementation approaches; incorporate new approaches or knowledge specific to working in Côte d’Ivoire; form or strengthen partnerships (if any); and develop strong concept notes. Note that competitors participating in co-creation are not required to partner or collaborate with each other.
The purpose of co-creation efforts are for competitors to:
The following activities will be part of co-creation:
USAID will make every effort to schedule the co-creation sessions at a reasonable time for all competitors and judges.
Concept Note: Submission Requirements
Based on the aforementioned co-creation activities, competitors who wish to continue competing for the grant may submit a concept note. USAID will share the concept note evaluation criteria at the beginning of the co-creation effort. The concept note is expected to be up to ten pages in length and will incorporate the following:
A. Updated intelligent forecasting model (one model), with revisions based on discussion and feedback provided by USAID;
B. Narrative describing the revised intelligent forecasting model, including a description of what changed; what didn’t; and a discussion on the reproducibility of the model in the field;
C. Revised and expanded narrative from the expression of interest which incorporates feedback from judges, USAID, and Ivorian stakeholders. This includes:
D. Narrative description of the approach and activities that will be undertaken as part of the grant;
E. Sustainability, a discussion of the challenges, and opportunities for the long term technical and financial sustainability of the model in the local public sector health system with minimal external support. Answers can also address how the model can be updated in light of new data and circumstances;
F. Scalability, a discussion of how the model could be scaled in the country, challenges and opportunities for scaling it to other country contexts, and any additional resources required for success;
G. List of deliverables and outputs; and
H. Proposed indicators to measure performance and proposed performance targets.
The following appendices should also be included and will not count towards the ten page limit:
The information provided on this website is not official U.S. Government information and does not represent the views or positions of the U.S. Agency for International Development or the U.S. Government.