Abstract
Forensic investigation is ushering into a new era of transformation propelled by rapid technological developments and innovations. The criminals are getting smarter, and crimes are becoming more complex; in such a time dissemination of justice requires commensurate technological enhancement. This chapter explores the vast potential of AI in revolutionizing Forensic Science and provides a succinct overview into the applicability of artificial intelligence (AI) and machine learning (ML) to facilitate classification, characterization, discrimination, differentiation, and recognition of forensic exhibits. This chapter further delves into the fundamental principles of supervised, unsupervised, semi-supervised, and reinforcement learning approaches and describes common ML methods which are frequently employed by researchers of this field.
TopBackground
Forensic Science, an interdisciplinary discipline, applies principles of various scientific branches to link individuals, locations and objects involved in criminal activities and aid in investigation and adjudication of civil and criminal cases (Houck, 2007). With increasing number and complexities of crime, growing awareness among criminals, transport revolution, weakening of social cohesion and faster dissemination of information, the importance of forensic science and criminalistics is increasing more and more in investigation and dissemination of justice. This domain requiring meticulous observation and keen analysis often falters in court of law because of human biasness and errors. As the twenty first century ushers into a digital world, inventions and innovations endowed with the capacity to swiftly analyse vast quantities of data and discern intricate patterns offer a comprehensive solution to interpret and solve complex criminal cases. In the labyrinth of forensic enquiry, Artificial Intelligence (AI), a burgeoning integration of human intelligence and machine ingenuity, promises to revolutionize traditional investigation methodologies and augment the capabilities of forensic experts and law enforcement agencies; ushering in an age marked by swifter processing, sharper insights, greater accuracy, higher precision and bias free results.
Key Terms in this Chapter
Semi-Supervised Learning: This is a hybrid approach combining supervised and unsupervised learning which uses both labelled and unlabeled datapoints in training algorithm.
Reinforcement Learning: It is a feedback-based learning mechanism, where machine automatically learns by itself based on the perception it receives from environment. The machine generates its future action based on the rewards and penalties.
Deep Learning: Deep learning is an advanced machine learning technique which employs multiple layers of non-linear information processing to learn representation of data with multiple levels of abstraction in order to model complex relationships among data.
Machine Learning: Machine learning concerns with the ways through which a machine acquires knowledge and gets trained to perform tasks without being explicitly programmed.
Artificial Intelligence: Artificial intelligence is defined as science behind imbuing computers and machines with the capability to simulate intellectual task akin to those performed by humans.
Supervised Learning: Supervised learning is a form of machine learning where the model is trained using data for which the correct label is already known.
Unsupervised Learning: Unsupervised learning seeks to discover new patterns of information in unlabeled raw input data which can then be utilized to assign new label to the data.