Systematic Poisoning Attack and Defense for Machine Learning in Healthcare
Project Abstract / Summary : p { margin-bottom: 0.25cm; direction: ltr; color: rgb(0, 0, 10); line-height: 120%; text-align: left; widows: 2; orphans: 2; }p.western { font-family: "Calibri",serif; font-size: 11pt; }p.cjk { font-family: "Calibri"; font-size: 11pt; }p.ctl { font-size: 11pt; }
Generally, doctors and medical professionals use their knowledge and experience to make decision for the diagnosis of diseases for patients. The huge amounts of patient’s medical records are stored in database of various hospitals which does not have much use. Main aim of this project is to develop a software application for immediate diagnose of a disease using these records which can be used by patients, interns in medical field or medical professionals. This system takes symptoms as an input, analyzes the symptoms and existing records and generates the corresponding disease.
Machine learning is being used in various application domains to discover patterns in large databases. And now a days in many healthcare and medical related applications, machine learning algorithms have been used to give optimal results. Healthcare related applications are often highly sensitive and, thus, any security breach would be catastrophic. Research has shown that some machine learning algorithms are vulnerable to attacks on their database. One such attack is Poisoning attack which augments the training sets with malicious data, leading to attack on the dataset. We are implementing an algorithm for systematic poisoning attack on medical database of medical diagnosis system. It can be applied on any machine learning algorithm and results in false diagnosis. A false diagnosis cause users to distrust the machine learning algorithm and even abandon the entire system and it makes false positive classifications which may cause distress.
To this attack algorithm we have presented a defense algorithm that is based on encryption technique. It identifies the malicious instances added into the medical database and immediately deletes it in order to avoid its impact on diagnosis. This defense algorithm does not makes any error and avoids any sort of poisoning attack.
Why did you choose to work on this project topic : Healthcare data is a sensitive and any malicious attack on it will compromise the health of an individual.This era of technology innovations where all the data is available online there is a chance that attackers attacking a medical database will be dangerous for the patient. Also if the its a Poisoning attack using Machine learning then detecting such an attack is even more difficult.So we have come up with a defense system which mitigates the effect of poisoning attack on a medical database.
Project Highlights : The current defense system of the medical database system only raises an alarm if the malicious instances get added in the database.We have Presented a system which identifies the malicious instances and immediately removes them from the database.
Project Category : CS / IT / Networking
------------------------------------------------------
Institute/College Name: St. John College of Engineering and Technology
City: Palghar
State: Maharashtra
Participating Team From: Final Year
Generally, doctors and medical professionals use their knowledge and experience to make decision for the diagnosis of diseases for patients. The huge amounts of patient’s medical records are stored in database of various hospitals which does not have much use. Main aim of this project is to develop a software application for immediate diagnose of a disease using these records which can be used by patients, interns in medical field or medical professionals. This system takes symptoms as an input, analyzes the symptoms and existing records and generates the corresponding disease.
Machine learning is being used in various application domains to discover patterns in large databases. And now a days in many healthcare and medical related applications, machine learning algorithms have been used to give optimal results. Healthcare related applications are often highly sensitive and, thus, any security breach would be catastrophic. Research has shown that some machine learning algorithms are vulnerable to attacks on their database. One such attack is Poisoning attack which augments the training sets with malicious data, leading to attack on the dataset. We are implementing an algorithm for systematic poisoning attack on medical database of medical diagnosis system. It can be applied on any machine learning algorithm and results in false diagnosis. A false diagnosis cause users to distrust the machine learning algorithm and even abandon the entire system and it makes false positive classifications which may cause distress.
To this attack algorithm we have presented a defense algorithm that is based on encryption technique. It identifies the malicious instances added into the medical database and immediately deletes it in order to avoid its impact on diagnosis. This defense algorithm does not makes any error and avoids any sort of poisoning attack.
Why did you choose to work on this project topic : Healthcare data is a sensitive and any malicious attack on it will compromise the health of an individual.This era of technology innovations where all the data is available online there is a chance that attackers attacking a medical database will be dangerous for the patient. Also if the its a Poisoning attack using Machine learning then detecting such an attack is even more difficult.So we have come up with a defense system which mitigates the effect of poisoning attack on a medical database.
Project Highlights : The current defense system of the medical database system only raises an alarm if the malicious instances get added in the database.We have Presented a system which identifies the malicious instances and immediately removes them from the database.
Project Category : CS / IT / Networking
------------------------------------------------------
Institute/College Name: St. John College of Engineering and Technology
City: Palghar
State: Maharashtra
Participating Team From: Final Year
Replies
You are reading an archived discussion.
Related Posts
Project Abstract / Summary : The use of QR code-based technologies and applications has become prevalent in recent years where QR codes are accepted to be a practical and intriguing...
Project Abstract / Summary : Advancement of technology has made our life simpler where everything is available at our fingertips. Controlling home appliances using an Android phone gives users the...
Project Abstract / Summary : The aim of the project is to enhance the efficiency of cooking fuel utility in the domestic field, in a secured way. It aims at...
Project Abstract / Summary : Main essencefor the thermal power plants that may be either coal or natural gas power plants to be in command of the temperature and pressure...
Project Abstract / Summary : In this era of depleting resources need of the hour is to preserve the environment. Hence, we have designed and fabricated a 4 wheeler 2...