10 Best Time Series projects you should practice (With Dataset)
Are you well versed with time series concepts? If yes, wonderful. Now, it’s time to put your knowledge into action and create some amazing projects. In this post, We’ll be discussing the 10 best time-series projects that will not just help you to build your portfolio in data science and machine learning but also help you to understand real-life problems and how can they leverage Data Science to make our life easier.
📹 YouTube | ☕️ BuyMeaCoffee | Linkedin
Table of Content
- Temperature Forecasting
- Daily Total Female Births in California Forecasting
- Web Traffic Time Series Forecasting
- Store Item Demand Forecasting
- Migration Prediction
- Air Quality Forecasting
- Amazon Stock Price Prediction
- Anomaly Detection
- Inflation Rate Forecasting
- ECG Anomaly Detection
Okay! so let’s discuss each project in detail and the reason why did we choose these.
1. Temperature Forecasting
Why Temperature Forecasting?
How wonderful is it if you could forecast the temperature of a city in advance! This could really help the government authorities as well as the general public to take appropriate action at right time. We all know that due to changing weather conditions, many people die every year. According to The Hindu, nearly 7,40,000 deaths are due to abnormally hot and cold temperatures related to climate change.
How to implement it?
Time Series Models like Autoregressive Integrated Moving Average (ARIMA), SARIMA, SARIMAX, etc. can help to help to forecast the temperature of upcoming days in advance. The biggest advantage is, they can also capture seasonal and cyclic trends.
You can find the dataset here.
2. Daily Total Female Births in California Forecasting
Why daily birth forecasting is important?
This project aims to forecast the number of female births. This can help the government to forecast the female births in their city/country. The government can create policies accordingly. According to Lancet Global Health, about 2,39,000 girls die each year in the country because of their gender.
Highly shocking! That’s why forecasting helps the government to prepare well to tackle these issues in the future.
How to implement it?
Again, Time Series models like ARIMA, AR, MA can help create models and get forecasting results.
The dataset contains the total number of female births recording in California, the USA during the year 1959.
You can find the dataset here.
3. Web Traffic Time Series Forecasting
Why Web Traffic Forecasting is important?
The project aims to forecast the amount of traffic on Wikipedia pages. This can help the website owners to forecast the traffic on their traffic. On the basis of that, they can take appropriate steps.
Suppose the traffic forecasted is very high, probably the website owners can think of putting load balancers to their website because if a website goes down even for a few minutes, it can incur a loss to the owner. In case of low traffic, they can probably think of new marketing strategies to drive traffic to their website.
How to implement it?
First, you need to web scrape the data from the Wikipedia pages, and then after pre-processing you can apply time series models like ARIMA and SARIMA. These methods are capable of capturing seasonal or cyclic trends as well.
You can find the dataset here.
4. Store Item Demand Forecasting
Why forecasting the sales in advance is important?
Just think how beneficial it is if you could forecast the sales in advance! From a business perspective, this is probably one of the most magical things that the store or a business can forecast the sales in advance. But, is this possible?
Yes, it is. Provided the past sales data, you can forecast the future sales in advance. Also, as we know that during the festive season, sales go high in many stores. So, we need to forecast the sales in the festive season too when the sales are at peak.
How to implement it?
The Time series models can capture the seasonal trends as well. SARIMA Model can be very useful for this scenario. They are able to capture seasonal trends also.
The dataset contains 5 years of store-item sales data and you need to predict 3 months of sales for 50 different items at 10 different stores. You can find the dataset here.
5. Migration Prediction
Why Migration Prediction is important?
The project aims to forecast the inflow of migrants into various European Countries. By doing so, the government authorities can be proactive in preparing to meet their needs and advocate for the political will to provide safe passage into Europe.
Assistance is needed to be provided to the migrants. That’s why forecasting is of prime importance.
You can find the dataset here.
6. Air Quality Forecasting
Why it is important?
According to WHO, around 4.2 million deaths occur every year across the globe due to respiratory diseases. Air pollution is a menace. It is very important to forecast the air quality. The government can take appropriate actions to tackle air pollution at an emergency level. This can help save a lot of lives.
How to implement it?
Time Series models like ARIMA, SARIMA, and SARIMAX can be used. Also, as we know that generally, pollution levels increase in winter. So, these models can identify the seasonal trends as well.
You can find the dataset here.
7. Amazon Stock Price Prediction
Why Stock Price Prediction?
Stock Market Prediction is the act of trying to determine the future value of company stock or other financial instrument traded on an exchange. The successful prediction of a stock’s future price can yield a high profit. Also, we know that the market is highly unstable and changes significantly when a new political party is formed after every 5 years due to a change in government policies for business.
The dataset contains the stock prices of Amazon for each day. You can find the dataset here.
8. Anomaly Detection
Why Anomaly Detection?
Not just forecasting, it is equally important to detect anomalies in the time series data. It is necessary to develop a diagnosis module to detect default in the systems. Achieving this can limit the consequences of failures that can be catastrophic for human goods and life. After detecting abnormality, the cause can be located and identified to make decisions.
How to implement it?
Anomaly Detection Algorithms like Isolation forest can be used. Also, deep learning techniques like LSTM and autoencoders can be used. ARIMA, SARIMA can also be used for forecasting purposes.
You can find the dataset here.
9. Inflation Rate Forecasting
Why Inflation Rate Forecasting is important?
Harmonized Index of consumer prices (HICP) is a consumer price index that is used to measure Inflation rates in Europe. It is used to implement the monetary policy. HICP is used to calculate the amount the average consumer would have to spend in a given year to buy the same basic goods and services. It also affects the livelihood of the people. A high inflation rate can be dangerous. That’s why forecasting them in advance can help the government to change any existing economic policies for the betterment of the economy of the nation.
You can find the dataset here.
10. ECG Anomaly Detection
Why it is an important use case?
The project aims to analyze real-time sensor data to identify any abnormal heartbeats. This is such a critical issue. Just imagine if the doctors can detect the abnormality, he/she can take appropriate action. This can save many lives and is the magic of Artificial Intelligence. AI can save humanity if deployed carefully.
You can find the dataset here.
Conclusion
Hope you were able to understand the importance of analyzing the time series data and what kind of problems it can solve. Time Series models can be harnessed to solve real-life use-cases.
If you enjoyed the blog, I kindly invite you to show your appreciation by giving it a clap on Medium. Your support means a lot and helps me continue sharing valuable content. Thank you! 👏
📹 YouTube | ☕️ BuyMeaCoffee | Linkedin
