The Random Forest Kernel and other kernels for big data from random partitions

We present Random Partition Kernels, a new class of kernels derived by demonstrating a natural connection between random partitions of objects and kernels between those objects. We show how the construction can be used to create kernels from methods that would not normally be viewed as random partitions, such as Random Forest. To demonstrate the potential of this method, we propose two new kernels, the Random Forest Kernel and the Fast Cluster Kernel, and show that these kernels consistently outperform standard kernels on problems involving real-world datasets. Finally, we show how the form of these kernels lend themselves to a natural approximation that is appropriate for certain big data problems, allowing $O(N)$ inference in methods such as Gaussian Processes, Support Vector Machines and Kernel PCA.

Thin film shape memory alloys and microactuators

For Micro-electro-mechanical System (MEMS) applications, TiNi-based thin film Shape Memory Alloys (SMAs) possess many desirable properties, such as high power density, large transformation stress and strain upon heating and cooling, superelasticity and biocompatibility. In this paper, recent development in TiNi-based thin film SMA and microactuator applications is discussed. The topics related to film deposition and characterisation is mainly focused on crystal nucleation and growth during annealing, film thickness effect, film texture, stress induced surface relief, wrinkling and trenches as well as Temperature Memory Effect (TME). The microactuator applications are mainly focused on microvalve and microcage for biological applications, micromirror for optical applications and data storage using nanoindentation method. Copyright © 2009, Inderscience Publishers.

Cryogenic single-shot spectroscopy of a floating poly-silicon gate transistor

Trapped electrons, located close to the channel of a transistor, are promising as data storage elements in non-classical information processing. Cryogenic microwave spectroscopy has shown that these electrons give rise to high quality factor resonances in the drain current and a post excitation dynamic behaviour that is related to the system lifetime. Using a floating poly-silicon gate transistor, single shot spectroscopy is performed to characterise the dynamic behaviour during excitation. This behaviour is seen to be dominated by the decay of the transient component, which gives rise to oscillations around the high quality factor resonance. © 2012 American Institute of Physics.

Kinetic modelling of growth and storage molecule production in microalgae under mixotrophic and autotrophic conditions.

In order to improve algal biofuel production on a commercial-scale, an understanding of algal growth and fuel molecule accumulation is essential. A mathematical model is presented that describes biomass growth and storage molecule (TAG lipid and starch) accumulation in the freshwater microalga Chlorella vulgaris, under mixotrophic and autotrophic conditions. Biomass growth was formulated based on the Droop model, while the storage molecule production was calculated based on the carbon balance within the algal cells incorporating carbon fixation via photosynthesis, organic carbon uptake and functional biomass growth. The model was validated with experimental growth data of C. vulgaris and was found to fit the data well. Sensitivity analysis showed that the model performance was highly sensitive to variations in parameters associated with nutrient factors, photosynthesis and light intensity. The maximum productivity and biomass concentration were achieved under mixotrophic nitrogen sufficient conditions, while the maximum storage content was obtained under mixotrophic nitrogen deficient conditions.

Kinetic modelling of growth and storage molecule production in microalgae under mixotrophic and autotrophic conditions

In order to improve algal biofuel production on a commercial-scale, an understanding of algal growth and fuel molecule accumulation is essential. A mathematical model is presented that describes biomass growth and storage molecule (TAG lipid and starch) accumulation in the freshwater microalga Chlorella vulgaris, under mixotrophic and autotrophic conditions. Biomass growth was formulated based on the Droop model, while the storage molecule production was calculated based on the carbon balance within the algal cells incorporating carbon fixation via photosynthesis, organic carbon uptake and functional biomass growth. The model was validated with experimental growth data of C. vulgaris and was found to fit the data well. Sensitivity analysis showed that the model performance was highly sensitive to variations in parameters associated with nutrient factors, photosynthesis and light intensity. The maximum productivity and biomass concentration were achieved under mixotrophic nitrogen sufficient conditions, while the maximum storage content was obtained under mixotrophic nitrogen deficient conditions. © 2014 Elsevier Ltd.