Project 4: West Nile Virus Prediction

Introduction

The West Nile Virus (WNV) has been a serious problem for the United States since 1999. The CDC has acknowledged it as the leading cause of mosquito-borne disease in the continental United States. However, there are no vaccines to prevent or medications to treat WNV in people -- according to the CDC, 1 in 5 people who are infected develop a fever and other symptoms, while 1 out of 150 infected people develop a serious, sometimes fatal, illness.

In Illinois, the West Nile virus was first identified in September 2001 when laboratory tests confirmed its presence in two dead crows found in the Chicago area. The following year, the state's first human cases and deaths from West Nile disease were recorded and all but two of the state's 102 counties eventually reported a positive human, bird, mosquito or horse. By the end of 2002, Illinois had counted more human cases (884) and deaths (64) than any other state in the United States.

Since then, Illinois and more specifically Chicago, has continued to suffer from multiple outbreaks of the West Nile Virus. From 2005 to 2016, a total of 1,371 human WNV cases were reported within Illinois. Out of these total reported cases, 906 cases (66%) were from the Chicago region (Cook and DuPage Counties).

With this in mind, our project is aimed at predicting outbreaks of the West Nile Virus. This will help the City of Chicago and Chicago Department of Public Health (CDPH) more efficiently and effectively allocate resources towards preventing transmission of this potentially deadly virus. Specifically, our model will use a combination of weather, time and location features to predict the presence of WNV within mosquito traps set up throughout Chicago.

Executive Summary

Our top-performing model was a Logistic Regression model (Lasso Regularization with an $\alpha$ of 1.0), which achieved an AUC score of 0.787 on Kaggle. This placed us among the top 15% of all teams that participated in the competition. To achieve these results, we used a range of feature engineering, selection and oversampling techniques including Synthetic Minority Over-Sampling (SMOTE). Some of our best predictors were temperature-related features, such as Tmax and WinterDepart. Time-based features like Week and Month were also crucial in helping our model to identify the presence of the West Nile Virus. We also saw other weather-related variables like humidlag4 play a role in our model. Location features like Longitude and trap location also played a minor role here.

Model Results (Sorted by Test AUC)

Model	Train AUC	Test AUC	Precision	Specificity	Recall	F-score
Logistic Regression	0.858627	0.823617	0.140553	0.750251	0.739394	0.236205
Ada Boosting	0.961605	0.808654	0.157986	0.837630	0.551515	0.245614
Extra Trees	0.974448	0.807757	0.148900	0.831604	0.533333	0.232804
Gradient Boosting	0.997995	0.801529	0.232673	0.948108	0.284848	0.256131
Random Forest	0.992065	0.797336	0.177616	0.886843	0.442424	0.253472
Support Vector Machine	0.938088	0.778879	0.140950	0.806160	0.575758	0.226460
Decision Tree	0.989197	0.724557	0.157360	0.888852	0.375758	0.221825

We also chose the Logistic Regression model due to its high recall score. Given that the West Nile Virus can lead to human death, it's imperative for false negatives to be minimized and for true positives to be maximized. Our Logistic Regression model has by far the best recall score out of all the other models (0.74) though it has weak precision and specificity scores, we believe that this is a fair trade-off as incorrectly predicting the lack of WNV can increase the chances of an outbreak, leading to potential snowball effects on hospitalization rates and the economy.

Recommendations

Our model has shown that certain areas are particularly 'dense' in terms of WNV-positive pools and pool proximity. In conjunction to this, our model also predicted several traps that have an 80% probability or greater of a WNV outbreak. We believe that the neighborhoods in which these traps are located should be an immediate focus for mosquito control efforts. These areas have been highlighted with a red circle below.

We've extrapolated that these are the neighborhoods that have a high risk of WNV:

• Elk Grove Village (7,500 acres)
• Des Plains (9,000 acres)
• Norridge (1,100 acres)
• Lincolnwood (1,700 acres)
• Stickney (1,200 acres)
• Forest View (900 acres)
• Morton Grove (3,100 acres)

Around 24,500 acres of area in Chicago were identified as high risk, housing an approximate population of 148,500 people.

We recommend using the following methods to deal with areas with a high WNV risk:

Automation with drones

Drones can serve multiple purposes when it comes to combatting WNV:

1. They can collect aerial images that can be analyzed and used to identify and map breeding sites—such as cisterns, pots and buckets, old tires and flower pots. These images can be aggregated into accurate maps to support targeted application of larvicides (insecticides that specifically target the larval stage of an insect) at these potential breeding sites.
2. Once larval habitats are identified, drones can be equipped to carry and apply larvicides and/or adulticides to small targeted areas. These drones can also be fitted with a global positioning system (GPS) that can track flight patterns in conjunction with insecticide application. An operator can remotely pilot the drone or, in some cases, autopilot programs may be available for pre-programmed flights. Drones can be useful to target specific areas with larvicides or adulticides, as an alternative to truck-mounted applications that may require a high degree of drift of droplets to reach a target area in remote locations. These drones can spray potentially up to 80 acres in a day's work.

In summary, drones could be more environmentally friendly than doing the same spraying procedure on foot, and are likely to be a lot more accurate due to the ability to spray from a fully-vertical angle.

Adoption of best practices

The city of Chicago can aim to reduce spraying target areas by adopting guidelines from tropical countries like Singapore that have proven to be successful in combatting mosquito-borne viruses. For example, this could involve active on-the-ground checks of homes and premises for mosquito habitats, where public officers provide advice on steps you can take to prevent mosquito breeding and impose penalties if premises are found with mosquito breeding.

Through automation and more efficient mosquito vector control processes, we can reduce the cost of spraying, enabling the city of Chicago to save human lives and prevent further outbreaks of the West Nile Virus.

Data Dictionary

Name	Dataset	Type	Description
Id	train/test	int	ID number of the record
Date	train/test	datetime	Date the WNV test was performed
Address	train/test	datetime	Approximate trap address retrieved from GeoCoder
Species	train/test	str	Mosquito species in trap
Block	train/test	str	Block Number of address
Street	train/test	str	Street of address
Trap	train/test	str	ID number of the trap
AddressNumberAndStreet	train/test	str	Approximate address retrieved from GeoCoder
Latitude	train/test	float	Latitude retrieved from GeoCoder
Longitude	train/test	float	Longitude retrieved from GeoCoder
AddressAccuracy	train/test	int	Accuracy of information returned from GeoCoder
NumMosquitos	train	int	Number of mosquitoes in a sample
WnvPresent	train	int	Whether or not WNV is present in a sample (1 = present; 0 = absent)
Station	weather	str	Station 1 or 2
Date	weather	datetime	Date of measurement (YY/MM/DD)
Tmax	weather	float	The highest temperature for the day in degrees Fahrenheit (F).
Tmin	weather	float	The lowest temperature for the day in degrees Fahrenheit (F).
Tavg	weather	float	The average temperature for the day in degrees Fahrenheit (F).
Depart	weather	float	Departure from normal temperature. The difference between column 4 and the 30 year normal temperature for this date. A minus (-) is number of degrees below normal. A zero (0) indicates that the average for that day was the normal temperature. (F)
DewPoint	weather	float	Average Dew Point temperature (F)
WetBulb	weather	float	Average Wet Bulb temperature (F)
Heat	weather	float	Heating Degree Days base 65F, season begins with July.
Cool	weather	float	Cooling Degree Days base 65F, season begins with January.
Sunrise	weather	float	Time of sunrise
Sunset	weather	float	Time of sunset
CodeSum	weather	str	Significant weather phenomena
Depth	weather	float	Snow depth on the ground to the nearest inch
Water1	weather	float	Water Equivalent in inches
SnowFall	weather	float	Total snowfall for the day to the nearest tenth of an inch.
PrecipTotal	weather	float	Total precipitation for the day to the nearest hundredth of an inch. This includes all forms of precipitation, both liquid and water equivalent of any snow or ice that occurred
StnPressure	weather	float	Average station pressure in hg (inches)
SeaLevel	weather	float	Average sea level pressure in hg (inches)
ResultSpeed	weather	float	Resultant Wind Speed - Vector sum of wind speeds divided by number of observations (MPH)
ResultDir	weather	float	Resultant Wind Direction - Vector sum of wind divided by number of observations (in tens of degrees)
AvgSpeed	weather	float	Average wind speed (MPH)
Date	spray	datetime	The date the pesticide was sprayed (YY/MM/DD)
Time	spray	datetime	Time of spray
Latitude	spray	float	The latitude of the location sprayed.
Longitude	spray	float	The longitude of the location sprayed.