Opportunities are available for Master’s thesis projects and Bachelor’s internships on theoretical or implementation-oriented topics in the area of Natural Computing and Unconventional Computing.

For further information, please contact the professor at claudio.zandron@unimib.it

 

Internal Thesis/Internship opportunities (university lab)

• Membrane systems are computing models inspired by the functioning of the cell. Two papers describing the model can be downloaded here and here. Various topics can be considered to be investigated in this framework, such as:
  • Cross-fertilization of this framework with other bioinspired computational models, such as Genetic Algorithms, Particle Swarm Optimization, ant computing, and others
  • Error correction in membrane systems
  • Computational complexity aspects: solutions for computationally hard problems, comparisons of complexity classes for membrane systems with standard complexity classes
  • Inferring features of membrane systems starting from their computations
  • Implementing algorithmic techniques in the framework of membrane systems, like, e.g., Fixed Parameter Tractability, or SAT solvers
• Spiking neural networks are a likely representation of the neuronal model. With the use of the Nengo framework, it is possible to develop spiking neural networks that open up possible directions:
  • Conversion of known neural network architectures into spiking networks: considering known neural network architectures, it is possible to convert them into spiking neural networks and then apply transfer learning techniques to refine the results. Once the transfer learning is completed, the advantages and disadvantages can be highlighted
  • Sequential networks: development and implementation of a sequential spiking neural network. Sequential networks read sequences of input, such as a series of numbers or a natural language sentence
  • Modelling of biological processes: Spiking neural networks are used to model cognitive processes. One possibility is to delve into some aspects of the book How to Build a Brain: A Neural Architecture for Biological Cognition and propose new features
  • Federated learning (also known as collaborative learning) is a machine learning technique that trains an algorithm via multiple independent sessions, each using its own dataset, in contrast to traditional centralized machine learning techniques. The work consists of implementing federated learning methods in the framework of Spiking neural membrane systems
• Water computing models are parallel computing systems without any central control, constituted by tanks containing water, valves to control the flow of water, and pipes connecting tanks. The flow of water is solely regulated by local measurements of tank filling levels in a finite number of water tanks, each capable of holding an initial volume of water and storing or collecting water up to a maximum capacity. Analysis of computing power and implementations of elements like perceptrons or neural networks are interesting aspects to be investigated. One possible approach is the one used in this work.

A more recent and detailed list of thesis ideas, proposed by PhD student Erba Sandro, is available here (Italian version here). For further information or questions regarding these opportunities, write to s.erba9@campus.unimib.it

External Thesis/Internship opportunities (companies)

30/03/25 Internship opportunities are available at the portal Cercaofficina.it.
Cercaofficina is a platform active since 2013 that allows motorists to compare personalized quotes from specialized workshops in their area of interest.
Technological evolution requires both technical and functional updates in some areas of the portal.
At the same time, a replacement of the historically used development framework (CakePHP) has begun, moving toward microservice-based technologies using APIs.
The intern will be assigned to one or more of the following projects:

• Programming path
  • A – Backend re-engineering of the “quote request funnel” of the motorists’ personal area of cercaofficina.it, and creation of an API layer for its functionalities
    (Technologies: JS, TypeScript, NestJS)
  • B – Analysis and adaptation/evolution of the algorithm for generating personalized and real-time estimates. The proprietary algorithm retrieves structured data from various sources (automotive labor times, motor vehicle registry, spare parts prices) and generates real-time estimates for motorists.
    This evolution is required to enable the algorithm to achieve greater accuracy in calculating estimates and a lower failure rate.
    (Technologies: JS, TypeScript, NestJS, PHP)
  • C – Frontend rewrite of the motorists’ personal area using JS/Vue.js. A specialized company has been tasked with designing a new user experience for the “estimates request flow” of the portal. The intern will join the frontend development team and actively participate in the development of the new interface.
    (Technologies: JS, TypeScript, Vue.js, Nuxt3)
• Database path
  • D – Analysis of Cercaofficina.it’s relational database (over 250 tables) to identify improvement proposals in terms of efficiency/scalability.
    This will be followed by re-engineering and modeling of certain database areas based on established best practices.
  • E – Creation of dynamic and real-time dashboards to provide company departments with business-related data (KPIs, data aggregations).
    The intern will assist in requirements analysis, data aggregation, and dashboard development.
    (Technologies: MySQL, tools such as Metabase, Cluvio).

 

---

Natural Computing Lab
Building U14 – Room 1046
Viale Sarca 336
20126 – MILANO
DISCo- Università Milano-Bicocca