The Challenge

How can you use machine learning to help your own community?

This Meridian Stories challenge encourages teams of three to four students to use their creativity, intellect, problem-solving skills, and cooperation to come up with a machine learning-based solution that can help to solve a community issue. Machine learning is defined as the science of getting computers to act without explicitly being told what to do by a person. Most often, machine learning is dependent on a computer reading and analyzing enormous amounts of data – an amount that humans wouldn’t be able to comprehend efficiently. By analyzing enormous amounts of data, a machine learning-based solution is able to predict trends in the future. In this competition, students will learn about machine learning in a capacity that will allow them to come up with real-world applications of machine learning.

The final product for this competition will be a 3 to 4 minute video demonstrating a problem in the community, proposing a realistic machine learning-based solution, and explaining the potential effects that the solution could have on the community. The format of the video is a pitch that students are sending to a community organization – one that they will need to identify or make up – to convince the organization to implement the solution. The solution should be realistic – plausible and economical – and should have no major disadvantages.

Specifically, your team needs to meet these three criteria in order to succeed:

1. Research the value of applying machine learning in communities. The best way to do this is by finding examples of other communities implementing machine learning to solve an issue.
2. Identify and define the problem. The problem will most likely be an issue of efficiency. Here are a few examples of issues that can be solved by machine learning:
   1. Outdated subway systems in large cities such as New York can be extremely inefficient. Often times, especially during rush hours, the accumulation of people waiting for a train will increase faster than the rate at which trains can take people to their destinations. This causes a bottleneck, which greatly lowers the efficiency of the entire city. Machine learning can help to combat this inefficiency by learning when and where bottlenecking could occur. Using the log of subway entry swipe cards, machine learning can determine when each station is busiest and allocate more trains to that station. For example, if Marcy Avenue station is extremely popular at 6PM but Halsey Avenue station is barren at the same time, then the machine learning solution would reallocate trains that would normally serve Halsey Avenue and send them to Marcy Avenue.
   2. Food can be a huge source of waste for organizations such as schools. Machine learning can help to cut down on waste by learning the exact types and quantities of food that are consumed. By knowing this information, machine learning can predict food trends and schools can buy exactly what they need without having surplus (or waste). For example, a machine learning solution could predict that the consumption of cold beverages such as juices and milk would rise substantially during Spring and Summer months. Using this knowledge, the school could increase the quantity of cold beverages sold during the Spring and Summer, and lower the quantities during the Fall and Winter.
3. Come up with a solution. Once a problem has been identified, come up with a machine learning-based solution that could help resolve the problem. This solution should be reasonable within the general means of your community. Keep in mind that this Challenge is not about actually applying machine learning – data gathering and analysis – but about finding a problem that the application of machine learning can help to solve, and then a) articulating that application; and b) predicting the outcome.
4. Present your gatherings from steps 1, 2, and 3 in a video pitch to an appropriate organization. This organization could be your school, city, state, etc.

Deliverables include:

• Learning Machine Learning Video (this is the only Meridian Stories deliverable)
• Summary of Problem and Proposed Machine Learning Application (at teacher’s discretion)
• Shooting Script (at teacher’s discretion)

 

Assumptions and Logistics

Time Frame – We recommend that this Meridian Stories Competition takes place inside of a three to four-week time frame.

Length – All Meridian Stories submissions should be under 4 minutes in length, unless otherwise specified.

 

Submissions – Keep in mind that each school can only submit three submissions per Competition (so while the entire class can participate in the Challenge, only three can be submitted to Meridian Stories for Mentor review and scoring).

Teacher Reviews – All reviews by the teacher are at the discretion of the teacher and all suggested paper deliverables are due only to the teacher. The only deliverable to Meridian Stories is the media work.

Teacher’s Role and Technology Integrator – While it is helpful to have a Technology Integrator involved, they are not usually necessary: the students already know how to produce the media. And if they don’t, part of their challenge is to figure it out. They will! The teacher’s primary function in these Challenges is to guide the students as they engage with the content.  You don’t need to know editing, sound design, shooting or storyboarding: you just need to know your content area.

Digital Rules/Literacy – We strongly recommend that all students follow the rules of Digital Citizenry in their proper usage and/or citation of images, music and text taken from other sources. This recommendation includes producing a citations page at the end of your entry, if applicable. See the Digital Rules area in the Meridian Resources section of the site for guidance.

We strongly recommend that students do not put their last names on the piece either at the start or finish, during the credits.

Location – Try not to shoot in a classroom at your school. The classroom, no matter how you dress it up, looks like a classroom and can negatively impact the story you are trying to tell.

 

Slate – All media work must begin with a slate that provides:

1. the title of the piece;
2. the name of the school submitting; and
3. the wording ‘Permission Granted’ which gives Meridian Stories the right to a) publicly display the submission in question on, as linked from or related to or in support of Meridian Stories digital media; and b) use it for educational purposes only.
4. We strongly recommend that students do not put their last names on the piece either at the start or finish, during the credits.

 

Collaboration – We strongly recommend that students work in teams of 3-4: part of the educational value is around building collaborative skill sets. But students may work individually.

Presentation – We strongly recommend that at the end of this process, the student teams present their work either to the class and/or to assembled parents and friends as a way to showcase their work. The workforce considers Presentational Skills to be a key asset and we encourage you to allow students to practice this skill set as often as possible. These short videos provide a great opportunity for kids to practice their public presentational skills.

Our research indicates this to be a really useful exercise for two additional reasons:

1. Students actually learn from their peers’ presentations – it is useful to hear a perspective that is not just the teacher’s; and
2. The public setting – painful as it is for some students – provides them with an opportunity to ‘own’ their work and to be more accountable.

 

Process

During Phase I, student teams will:

• Research how machine learning has been implemented by other communities. A great way to do this is by looking for online articles or videos of examples. You should organize your examples in a way that is easy to understand. Generally, you should be able to break your examples down into three parts: a problem, a machine learning-based solution, and the effects of the solution.
• Research a local issue that can be solved by machine learning. This issue can range in severity from something that is simply bothersome to something that is a pressing need. Be sure to spend part of the video detailing the issue and explaining why you chose it.
  • Teacher’s Option: Problem and Proposed Machine Learning Application – Teachers may require that teams hand in a two page summary outlining the problem they have chosen to solve, with support as to why Machine Learning may represent the best approach.
• Come up with a machine learning-based solution that can help to solve the identified issue. Simplified, a machine learning-based solution is essentially just data analysis. For this step, it is critical to consider these three things:
  • Machine learning-based solutions always rely on data. In the New York subway example, the data was the log of entry card swipes at each station.
  • The data must be Data is useless unless it is analyzed. For example, in the New York subway example, data shows us what stations are being used and at what time. However, we must take this data one step further if we want to use it. By analyzing this data, we are able find out which stations are busiest at certain times.
  • There must be a method of implementing the results of the data analysis, usually by predicting trends. In the New York subway example, data analysis revealed that some stations were busier than others at certain times. Using these results, the method was predicting which stations would be busy at what hours and allocating trains to those stations to decrease the bottlenecking of passengers.
  • The machine learning-based solution that you come up with must rely on community data that can be analyzed, and a method of implementing the analysis to solve the issue must exist. Include these three points in your video.

To clarify, here are the examples laid out in a chart:

Example	Data	Analysis	Method
New York Subway	Log of card swipes to get into subway stations	Determining which stations are busiest and at what times	Using analysis to allocate more trains to stations during when they are busiest
Cafeteria Food	The amount of food that is consumed by students	Determining which foods are most popular and when they most consumed	Using analysis to buy the correct amount of food necessary to reduce waste

 

• Brainstorm the effects that you think the solution will have on the community. In what ways will your community benefit? Are there any possible negative impacts that the machine learning solution could have? This step is straightforward.

 

During Phase II, student teams will:

• Create the organization to whom you will be presenting this pitch. The more you know about this audience and what they want, the better able you are to customize your video to them.
• Brainstorm about the key ideas that will inform your pitch. Here are some questions and ideas to consider:
• What is the most persuasive and potent idea in your research: the problem, the machine learning, problem-solving application or the potential benefits? Where, in this pitch, does this most potent idea go? At the beginning, or as a climactic ending?
• What are the two or three chief ‘selling points’ for your pitch? These points could form the spine of your pitch. Keep in mind that your goal is convince the organization (that you have invented) to invest in your solution; to back it and then finance it. So, persuasion is a part of the story that you need to tell.
• Who are the voices/characters that you want involved in this pitch? Sales people? Scientists? Possible end users? Will there be interviews?
  • Interviews with local people could be a key aspect of this for you. In order to really understand the problem, for example, you need to talk to people that are affected by it. Community engagement in the form of strategic interviews could be critical to the success of your video.
• Watch commercials on TV. You may want to pay special attention to the shopping channels that specialize in selling techniques. What works and what doesn’t work? Are there any ideas that you can adapt to your investor pitch?
• Your team should now have a) the problem, machine learning application to solve the problem, and the predicted outcomes; and b) a handful of creative ideas from your brainstorm above about how to position all of this content. Decide on the format or approach to your investor pitch and create a script outline.
• Draft the script.
• Discuss and map out the imagery needed to tell your story. Oftentimes a storyboard is the best process for this.
• Pre-produce the pitch:
• Scout locations for shooting (if this is being shot on location);
• Contact the people that you will need to include;
• Research, as necessary, the still images that you will integrate into your pitch;
• Create costumes, props and other set pieces, as needed;
• Prepare the logistics for the actual shooting of the pitch; and
• Rehearse the scenes that will comprise the pitch.

 

During Phase III, student teams will:

1. Finalize the script
   1. Teacher’s Option: Shooting Script – Teachers may require that teams hand in their Shooting Script
2. Shoot the video
3. Record voice overs or narration if necessary
4. Edit the video
5. Post-produce the video, adding sound and effects if needed

 

Media Support Resources

Meridian Stories provides two forms of support for the student teams: 1.    Media Innovators and Artists – This is a series of three to four-minute videos featuring artists and innovative professionals who offer important advice, specifically for Meridian Stories, in the areas of creativity and production. 2.    Meridian Tips – These are short documents that offer student teams key tips in the areas of creativity, production, game design and digital citizenry. Recommended review, as a team, for this Competition include:
Meridian Innovators and Artists	Media Resource Collection
On Documentary Films – Sarah Childress Interviewing Techniques – Tom Pierce On Producing – Tom Pierce On Editing – Tom Pierce	“Conducting an Interview” “Sound Recording Basics” “Six Principles of Documentary Filmmaking” “Video Editing Basics”

 

 

Evaluation Rubric – Learning Machine Learning

CONTENT COMMAND
Criteria	1 – 3	4 – 7	8 – 10
Problem Selection	The selected problem is not well defined and/or does not address a local need	The problem and the need to address it are evident	The problem is clearly defined and addresses a clear local need
Developing Possible Solutions	The machine learning-based solution is not possible to implement or does not solve the problem	The machine learning- based solution is possible to implement and it may solve the problem, but is not practical for the community or organization	The machine learning based solution is inventive, practical, economical, and has a high probability of working
Presenting Potential Effects of the Solution	The potential effects of the solution are not presented	The potential effects of the solution are presented	The potential effects of the solution are clearly and thoroughly presented.
STEAM Implementation	The project demonstrates weak utilization and little understanding of what machine learning is and how it can be implemented in the real-world.	The project reveals sufficient utilization and understanding of what machine learning is and how it can be implemented in the real-world.	The project reveals a thoughtful utilization and thorough understanding of what machine learning is and how it can be implemented in the real-world.
STORYTELLING COMMAND
Criteria	1 – 3	4 – 7	8 – 10
Scripting	The script does not convey the content in a well-organized or consistently engaging fashion	The script adequately conveys the content	The script clearly conveys the content in an engaging and compelling narrative
Persuasion	The pitch does not successfully persuade	The pitch is inconsistently persuasive and engaging	The pitch is consistently persuasive and engaging
MEDIA COMMAND
Criteria	1 – 3	4 – 7	8 – 10
Mixed Visual Media	The use of video, stills, animation, graphics and/or text was often confusing and not well matched to the goals of the video	The use of video, stills, animation, graphics and/or text was suitable to the goals of the video	The use of video, stills, animation, graphics and/or text was engaging, visually interesting and well matched to the goals of the video
Sound Design	The mix of music and sound did not enhance elements of the video	The mix of music and sound enhanced the video	The mix of music and sound greatly enhanced the video
Editing	The video feels patched together	The visual editing of the video is fluid	The visual editing is fluid and creative, resulting in an engaging video experience

 

21ST CENTURY SKILLS COMMAND (for teachers only)
Criteria	1-3	4-7	8-10
Collaborative Thinking	The group did not work together effectively and/or did not share the work equally	The group worked together effectively and had no major issues	The group demonstrated flexibility in making compromises and valued the contributions of each group member
Creativity and Innovation	The group did not make a solid effort to create anything new or innovative	The group was able to brainstorm new and inventive ideas, but was inconsistent in their evaluation and implementation of those ideas	The group brainstormed many inventive ideas and was able to evaluate, refine and implement them effectively
Initiative and Self-Direction	The group was unable to set attainable goals, work independently and manage their time effectively	The group required some additional help, but was able to complete the project on time with few problems	The group set attainable goals, worked independently and managed their time effectively, demonstrating a disciplined commitment to the project

 

Essential Questions

1. What is machine learning and how is it a relevant problem-solving process locally?
2. What are some local issues in your community that could benefit from a machine learning approach to problem solving?
3. How does one research and organize information in a way that leads to informed scientific speculation?
4. How does data serve as a tool to make our communities better?
5. How has immersion in the creation of original content and the production of digital media – exercising one’s creativity, critical thinking, and digital literacy skills – deepened the overall educational experience?
6. How has working on a team – practicing one’s collaborative skills – changed the learning experience?

 

Student Proficiencies

1. The student will gain a better understanding of how machine learning is implemented in the real world to solve issues and make communities more efficient.
2. The student will gain an increased awareness and understanding of some of the needs, opportunities and constraints of their community.
3. The student will understand the processes involved with researching a topic, synthesizing research, organizing research, and applying research to form a cohesive final product.
4. The student will understand that data and data collection are integral in solving issues that exceed the capabilities of the human mind.
5. The student will utilize 21^st century skills, with a focus on creativity, critical thinking and digital literacy, in their process of translating scientific content – the data collected – into a persuasive narrative.
6. The student will have an increased awareness of the challenges and rewards of team collaboration. Collaboration – the ability to work with others – is considered one of the most important 21^st century skills to develop in students as they prepare for life after secondary school.

 

Common Core and Next Generation Science Standards (NGSS) Curricular Correlations

The Learning Machine Learning Competition addresses a range of curricular objectives that are articulated in the NGSS and Common Core Mathematics Standards. Below please find the standards that are addressed, either wholly or in part.

Common Core – Mathematics

1. MATH.CONTENT.HSS.IC.A.1 – Understand statistics as a process for making inferences about population parameters based on a random sample from that population.
2. MATH.CONTENT.HSS.IC.B.6 – Evaluate reports based on data.
3. High School Modeling Standard

 

 

 

 

NGSS Standards

4. MS-ETS1-1 Engineering Design – Define the criteria and constraints of a design problem with sufficient precision to ensure a successful solution, taking into account relevant scientific principles and potential impacts on people and the natural environment that may limit possible solutions.
5. MS-ETS1-2 Engineering Design – Evaluate competing design solutions using a systematic process to determine how well they meet the criteria and constraints of the problem.
6. MS-ETS1-3 Engineering Design – Analyze data from tests to determine similarities and differences among several design solutions to identify the best characteristics of each that can be combined into a new solution to better meet the criteria for success.
7. MS-ETS1-4 Engineering Design – Develop a model to generate data for iterative testing and modification of a proposed object, tool, or process such that an optimal design can be achieved.
8. HS-ETS1-1 Engineering Design – Analyze a major global challenge to specify qualitative and quantitative criteria and constraints for solutions that account for societal needs and wants.
9. HS-ETS1-2 Engineering Design – Design a solution to a complex real-world problem by breaking it down into smaller, more manageable problems that can be solved through engineering.
10. HS-ETS1-3 Engineering Design – Evaluate a solution to a complex real-world problem based on prioritized criteria and trade-offs that account for a range of constraints, including cost, safety, reliability, and aesthetics as well as possible social, cultural, and environmental impacts.