5. Conclusions and Recommendations
We are a consulting firm called Quantum Analytics. Using data collection and analysis of seismic movements in the United States, Japan, and Chile, our consultancy utilizes a specific machine learning model to classify seismic movements and detect patterns. By validating the accuracy of our model, we can ensure that it provides reliable results to inform citizens about potential seismic hazards. Finally, we effectively communicate these results to reach as many people as possible and contribute to public safety.
Unpredictable natural events such as earthquakes claim the lives of hundreds of people when they occur. They can also cause secondary events such as tsunamis, volcano eruptions, nuclear hazards, and the condition of structures after these events is unknown. Current alerts using numbers or colors do not provide much information about secondary events or the destructiveness of an earthquake. We are faced with the need to communicate and alert the population of a region about the possibility of natural events, their destructiveness, and their consequences.
This team proposes an analysis of the situation in the last years 2018-2023 and a method of earthquake classification.
Go back to top
The scope of the project will be for the countries: Chile, United States, and Japan. Data from their respective observatories will be used. The use of external data or data from other countries is not ruled out to improve the model. The model will be a classification model.
- KPI: Alert Response Time:
This measures the time it takes for the alert system to notify the people after an earthquake event occurs. This KPI helps assess the speed and efficiency of your alert.
- KPI: Alert Reach:
Assess the percentage of the population reached by the alerts. This KPI provides insights into the coverage and penetration of the alert system and helps identify areas or demographic groups that may need additional attention.
- KPI: Impact Assesment:
Assess the impact of earthquakes by analyzing their relationships with secondary events, such as tsunamis and volcanic hazards. This KPI can provide insights into the potential consequences. Increase accuracy and clasification quality.
All KPIs will be measured annually.
The proposed solution is an alert API that can be implemented on various social media platforms and mobile applications. To achieve this, the following tasks are proposed.
Go back to top
Data sources:
Map of involved countries.
Go back to top
Data extraction is carried out from different information sources on the web, some data was extracted from an API in JASON format, in other cases a WEB SCRAPING of web pages was carried out, which will be processed directly as required by the MACHINE LEARNING Model and for a more detailed analysis.
In this stage of the project, the columns are established and the labels are adapted with the necessary data required by the classification model. The data to be reviewed by the data analyst is also considered. Functions of data overwriting and upload to the RDS server are developed.
For the classification model, we choose the model called 'Kmeans', it is an unsupervised model based on clustering and. That means the data will be grouped based on feature relationships, similarities, and does not need correct labeled data to train it.
Please take a look at our StreamLit dashboard.
Leveraging the powerful capabilities of streamlit as our core visualization tool, we have successfully generated a series of comprehensive and insightful graphs that shed light on the intricate nature of seismic activity in Japan, Chile, and the United States. These meticulously crafted visual representations serve as pivotal instruments in unraveling the complex dynamics of seismic movements, enabling us to discern patterns, trends, and potential risks with enhanced clarity. Our diligent efforts in harnessing the potential of streamlit have unlocked a wealth of previously inaccessible data, empowering us to conduct a thorough analysis of seismic phenomena. We cordially invite you to embark on an immersive journey through our detailed documentation, where we delve into the intricacies of our visualization process and its profound impact on seismic analysis.
Go back to top
Quantum Analytics utilizes data analysis and machine learning models to classify and detect patterns in seismic movements in the United States, Japan, and Chile. Our focus on validating the accuracy of the model ensures reliable results to inform citizens about potential seismic hazards. Through data visualization and effective communication, we contribute to public safety by reaching a wide audience. Our graphs showcase the sum of seismic activities, the magnitude-time distribution, and the intervals between consecutive seismic events, providing valuable insights to understand seismic activity and assist in risk assessment and preparedness efforts.
The graphs presented in our interactive StreamLit dashboard analysis are key tools for understanding seismic activity in Japan, Chile, and the United States. The "Sum of Seismic Activities per Country" graph allows for the identification of areas with higher seismic activity and assesses the intensity of earthquakes recorded in each country over time. The "Seismic Magnitude-Time Distribution" reveals temporal patterns and trends in seismic magnitudes, providing valuable information for risk assessment and planning of mitigation measures. The "Time Interval Between Consecutive Seisms" provides insights into the temporal patterns of earthquake occurrence, helping to understand clustering and gaps between seismic events. Lastly, the "Magnitude-Frequency Relationship" reveals the distribution of earthquakes according to their magnitudes, allowing for the identification of the frequency of events of different sizes. These combined graphs provide a comprehensive view of seismic activity, supporting informed decision-making and the implementation of appropriate safety measures.
Go back to top
Notion
Discord
Whatsapp
GitHub
Python
Numpy
Seaborn
Amazon AWS
Seaborn
Matplotlib
Streamlit
Python
Sklearn
FastAPI
Render
Go back to top
Go back to top
Dear users,
We would like to extend our heartfelt gratitude to each and every one of you who have come across our GitHub repository. Your presence and interest mean a lot to us.
As a dedicated and passionate team, we strive to deliver high-quality projects that showcase our love for technology and our commitment to constant improvement. We not only work hard to excel in the field of data, but we also aspire to become better individuals in every aspect of our lives.
Your support and engagement motivate us to push our boundaries and explore new horizons in the world of data. We deeply appreciate your time and attention, and we hope that our projects have provided value, insights, and inspiration to you.
We believe in the power of collaboration and knowledge sharing, and we are grateful for the opportunity to connect with like-minded individuals like you through this platform. Together, we can create a community that fosters growth, innovation, and positive change.
Once again, thank you for being a part of our journey. We are honored to have you here, and we look forward to continuing this exciting adventure with you.
With warm regards,
Quantum Analitycs Team
Go back to top