Gittins indices and constrained allocation in clinical trials

We explore the use of Gittins indices to search for near optimality in sequential clinical trials. Some adaptive allocation rules are proposed to achieve the following two objectives as far as possible: (i) to reduce the expected successes lost, (ii) to minimize the error probability at the end. Simulation results indicate the merits of the rules based on Gittins indices for small trial sizes. The rules are generalized to the case when neither of the response densities is known. Asymptotic optimality is derived for the constrained rules. A simple allocation rule is recommended for one-stage models. The simulation results indicate that it works better than both equal allocation and Bather's randomized allocation. We conclude with a discussion of possible further developments.

Traffic safety at road–rail level crossings using a driving simulator and traffic simulation

Several intelligent transportation systems (ITS) were used with an advanced driving simulator to assess its influence on driving behavior. Three types of ITS interventions were tested: video in vehicle, audio in vehicle, and on-road flashing marker. The results from the driving simulator were inputs for a developed model that used traffic microsimulation (VISSIM 5.4) to assess the safety interventions. Using a driving simulator, 58 participants were required to drive through active and passive crossings with and without an ITS device and in the presence or absence of an approaching train. The effect of changes in driver speed and compliance rate was greater at passive crossings than at active crossings. The slight difference in speed of drivers approaching ITS devices indicated that ITS helped drivers encounter crossings in a safer way. Since the traffic simulation was not able to replicate a dynamic speed change or a probability of stopping that varied depending on ITS safety devices, some modifications were made to the traffic simulation. The results showed that exposure to ITS devices at active crossings did not influence drivers’ behavior significantly according to the traffic performance indicator, such as delay time, number of stops, speed, and stopped delay. However, the results of traffic simulation for passive crossings, where low traffic volumes and low train headway normally occur, showed that ITS devices improved overall traffic performance.

Optimal eradication: when to stop looking for an invasive plant

The notion of being sure that you have completely eradicated an invasive species is fanciful because of imperfect detection and persistent seed banks. Eradication is commonly declared either on an ad hoc basis, on notions of seed bank longevity, or on setting arbitrary thresholds of 1% or 5% confidence that the species is not present. Rather than declaring eradication at some arbitrary level of confidence, we take an economic approach in which we stop looking when the expected costs outweigh the expected benefits. We develop theory that determines the number of years of absent surveys required to minimize the net expected cost. Given detection of a species is imperfect, the optimal stopping time is a trade-off between the cost of continued surveying and the cost of escape and damage if eradication is declared too soon. A simple rule of thumb compares well to the exact optimal solution using stochastic dynamic programming. Application of the approach to the eradication programme of Helenium amarum reveals that the actual stopping time was a precautionary one given the ranges for each parameter.

The Politics of Surveillance and Response to Disease Outbreaks: The New Frontier for States and Non-state Actors

The capacity to conduct international disease outbreak surveillance and share information about outbreaks quickly has empowered both State and Non-State Actors to take an active role in stopping the spread of disease by generating new technical means to identify potential pandemics through the creation of shared reporting platforms. Despite all the rhetoric about the importance of infectious disease surveillance, the concept itself has received relatively little critical attention from academics, practitioners, and policymakers. This book asks leading contributors in the field to engage with five key issues attached to international disease outbreak surveillance - transparency, local engagement, practical needs, integration, and appeal - to illuminate the political effect of these technologies on those who use surveillance, those who respond to surveillance, and those being monitored.

The politics of disease surveillance

The capacity to conduct international disease outbreak surveillance and share information about outbreaks quickly has empowered both State and Non-State Actors to take an active role in stopping the spread of disease by generating new technical means to identify potential pandemics through the creation of shared reporting platforms. Despite all the rhetoric about the importance of infectious disease surveillance, the concept itself has received relatively little critical attention from academics, practitioners, and policymakers. This book asks leading contributors in the field to engage with five key issues attached to international disease outbreak surveillance - transparency, local engagement, practical needs, integration, and appeal - to illuminate the political effect of these technologies on those who use surveillance, those who respond to surveillance, and those being monitored.

Thermal properties of single-phase Y2SiO5

Y2SiO5 is a promising candidate for oxidation-resistant or environmental/thermal barrier coatings (ETBC) due to its excellent high-temperature stability, low elastic modulus and low oxygen permeability. In this paper, we investigated the thermal properties of Y2SiO5 comprehensively, including thermal expansion, thermal diffusivity, heat capacity and thermal conductivity. It is interesting that Y2SiO5 has a very low thermal conductivity (∼1.40 W/m K) but a relatively high linear thermal expansion coefficient ((8.36 ± 0.5) × 10-6 K-1), suggesting compatible thermal and mechanical properties to some non-oxide ceramics and nickel superalloys as ETBC layer. Y2SiO5 is also an ideal EBC on YSZ TBC layer due to their close thermal expansion coefficients. As a continuous source of Y3+, it is predicted that Y2SiO5 EBC may prolong the lifetime of zirconia-based TBC by stopping the degradation aroused by the loss of Y stabilizer.

R&D in clean technology: A project choice model with learning

In this study, we investigate the qualitative and quantitative effects of an R&D subsidy for a clean technology and a Pigouvian tax on a dirty technology on environmental R&D when it is uncertain how long the research takes to complete. The model is formulated as an optimal stopping problem, in which the number of successes required to complete the R&D project is finite and learning about the probability of success is incorporated. We show that the optimal R&D subsidy with the consideration of learning is higher than that without it. We also find that an R&D subsidy performs better than a Pigouvian tax unless suppliers have sufficient incentives to continue cost-reduction efforts after the new technology success-fully replaces the old one. Moreover, by using a two-project model, we show that a uniform subsidy is better than a selective subsidy.

Parametrized actor-critic algorithms for finite-horizon MDPs

Due to their non-stationarity, finite-horizon Markov decision processes (FH-MDPs) have one probability transition matrix per stage. Thus the curse of dimensionality affects FH-MDPs more severely than infinite-horizon MDPs. We propose two parametrized 'actor-critic' algorithms to compute optimal policies for FH-MDPs. Both algorithms use the two-timescale stochastic approximation technique, thus simultaneously performing gradient search in the parametrized policy space (the 'actor') on a slower timescale and learning the policy gradient (the 'critic') via a faster recursion. This is in contrast to methods where critic recursions learn the cost-to-go proper. We show w.p 1 convergence to a set with the necessary condition for constrained optima. The proposed parameterization is for FHMDPs with compact action sets, although certain exceptions can be handled. Further, a third algorithm for stochastic control of stopping time processes is presented. We explain why current policy evaluation methods do not work as critic to the proposed actor recursion. Simulation results from flow-control in communication networks attest to the performance advantages of all three algorithms.

Molecular dynamics studies of nanoparticle impacts

This thesis concerns the dynamics of nanoparticle impacts on solid surfaces. These impacts occur, for instance, in space, where micro- and nanometeoroids hit surfaces of planets, moons, and spacecraft. On Earth, materials are bombarded with nanoparticles in cluster ion beam devices, in order to clean or smooth their surfaces, or to analyse their elemental composition. In both cases, the result depends on the combined effects of countless single impacts. However, the dynamics of single impacts must be understood before the overall effects of nanoparticle radiation can be modelled. In addition to applications, nanoparticle impacts are also important to basic research in the nanoscience field, because the impacts provide an excellent case to test the applicability of atomic-level interaction models to very dynamic conditions. In this thesis, the stopping of nanoparticles in matter is explored using classical molecular dynamics computer simulations. The materials investigated are gold, silicon, and silica. Impacts on silicon through a native oxide layer and formation of complex craters are also simulated. Nanoparticles up to a diameter of 20 nm (315000 atoms) were used as projectiles. The molecular dynamics method and interatomic potentials for silicon and gold are examined in this thesis. It is shown that the displacement cascade expansionmechanism and crater crown formation are very sensitive to the choice of atomic interaction model. However, the best of the current interatomic models can be utilized in nanoparticle impact simulation, if caution is exercised. The stopping of monatomic ions in matter is understood very well nowadays. However, interactions become very complex when several atoms impact on a surface simultaneously and within a short distance, as happens in a nanoparticle impact. A high energy density is deposited in a relatively small volume, which induces ejection of material and formation of a crater. Very high yields of excavated material are observed experimentally. In addition, the yields scale nonlinearly with the cluster size and impact energy at small cluster sizes, whereas in macroscopic hypervelocity impacts, the scaling 2 is linear. The aim of this thesis is to explore the atomistic mechanisms behind the nonlinear scaling at small cluster sizes. It is shown here that the nonlinear scaling of ejected material yield disappears at large impactor sizes because the stopping mechanism of nanoparticles gradually changes to the same mechanism as in macroscopic hypervelocity impacts. The high yields at small impactor size are due to the early escape of energetic atoms from the hot region. In addition, the sputtering yield is shown to depend very much on the spatial initial energy and momentum distributions that the nanoparticle induces in the material in the first phase of the impact. At the later phases, the ejection of material occurs by several mechanisms. The most important mechanism at high energies or at large cluster sizes is atomic cluster ejection from the transient liquid crown that surrounds the crater. The cluster impact dynamics detected in the simulations are in agreement with several recent experimental results. In addition, it is shown that relatively weak impacts can induce modifications on the surface of an amorphous target over a larger area than was previously expected. This is a probable explanation for the formation of the complex crater shapes observed on these surfaces with atomic force microscopy. Clusters that consist of hundreds of thousands of atoms induce long-range modifications in crystalline gold.

Generalization of Dynkin's Identity and Some Applications

Let X(t) be a right continuous temporally homogeneous Markov pro- cess, Tt the corresponding semigroup and A the weak infinitesimal genera- tor. Let g(t) be absolutely continuous and r a stopping time satisfying E.( S f I g(t) I dt) < oo and E.( f " I g'(t) I dt) < oo Then for f e 9iJ(A) with f(X(t)) right continuous the identity Exg(r)f(X(z)) - g(O)f(x) = E( 5 " g'(s)f(X(s)) ds) + E.( 5 " g(s)Af(X(s)) ds) is a simple generalization of Dynkin's identity (g(t) 1). With further restrictions on f and r the following identity is obtained as a corollary: Ex(f(X(z))) = f(x) + k! Ex~rkAkf(X(z))) + n-1E + (n ) )!.E,(so un-1Anf(X(u)) du). These identities are applied to processes with stationary independent increments to obtain a number of new and known results relating the moments of stopping times to the moments of the stopped processes.

NK-and T-cell function in chronic myeloid leukemia patients treated with interferon-? monotherapy

Chronic myeloid leukemia (CML) is one of the most studied human malignancies. It is caused by an autonomously active tyrosine kinase BCR-ABL, which is a result from a translocation between chromosomes 9 and 22 in the hematopoietic stem cell. As an outcome, a Philadelphia (Ph) chromosome is formed. BCR-ABL causes disturbed cell proliferation among other things. Although targeted tyrosine kinase inhibitor therapy has been developed in the beginning of the millenium and the survival rate has increased significantly, it is still not known why some patients benefit more from the treatment than others. Furthermore, the therapy is not considered to be curative. Before the era of tyrosine kinase inhibitors, the first-line treatment for CML was interferon-? (IFN-?). However, only a small proportion of patients benefitted from the treatment. Of these patients, a few were able to discontinue the treatment without renewal of the disease. The mechanism of IFN-? is not completely understood, but it is believed that differences in the immune system can be one of the reasons why some patients have better therapy response. Kreutzman, Rohon et al. have recently discovered that patients who have been able to stop IFN-? treatment have an increased number of NK- and T-cells. They also have a unique clonal T-cell population and more cytotoxic CD8+ T-cells and less CD4+ T-cells. The aim of this master’s thesis was to study the function of T- and NK-cells in IFN-? treated patients. Although it was shown earlier that IFN-? treated patients have increased NK-cell count, the function of these cells was unknown. Therefore, we have now investigated the killing potential of patients’ NK-cells, their activation status and cell surface antigen expression. In addition, we have also studied the activation status of patients’ T-cells and their cytotoxic properties. We observed that NK-cells from patients treated with IFN-? are unable to kill leukemic cells (K562) than NK-cells from healthy controls. In addition, patients on IFN-? treatment have more active T-cells and their NK-cells have an undifferentiated immunoregulatory phenotype. Patients that have been able to stop the treatment have anergic T-and NK-cells. As a conclusion our results suggest that IFN-? therapy induces increased NK-cell count, NK-cell immunoregulatory functions and more active T-cells. After stopping IFN-? therapy, NK- and T-cells from CML patients restore anergy typical for CML.

A quasi-steady performance prediction model for dynamic characteristics of a volute pump

A performance prediction model generally applicable for volute-type centrifugal pumps has been extended to predict the dynamic characteristics of a pump during its normal starting and stopping periods. Experiments have been conducted on a volute pump with different valve openings to study the dynamic behaviour of the pump during normal start-up and stopping, when a small length of discharge pipeline is connected to the discharge flange of the pump. Such experiments have also been conducted when the test pump was part of a hydraulic system, an experimental rig, where it is pumping against three similar pumps, known as supply pumps, connected in series, with the supply pumps kept idle or running. Instantaneous rotational speed, flowrate, and delivery and suction pressures of the pump were recorded and it was observed in all the tested cases that the change of pump behaviour during the transient period was quasi-steady, which validates the quasi-steady approach presented in this paper. The nature of variation of parameters during the transients has been discussed. The model-predicted dynamic head-capacity curves agree well with the experimental data for almost all the tested cases.

Centrifugal pump transient characteristics and analysis using the method of characteristics

A theoretical and experimental study has been carried out on the transient characteristics of a centrifugal pump during starting and stopping periods. Experiments have been conducted on a volute pump with different valve openings to study the dynamic behaviour of the pump during normal start up and stopping, when a small length of discharge pipe line is connected to discharge flange of the pump. Similar experiments have also been conducted when the test pump was part of a hydraulic system to study the system effect on the transient characteristics. Instantaneous rotational speed, flowrate, and delivery and suction pressures of the pump are recorded and it is observed in ail the tested cases that the change of pump behaviour during the transient period is quasi-steady. The dynamic characteristics of the pump have been analysed by a numerical model using the method of characteristics. The model is presented and the results are compared with the experimental data. As the model contains speed acceleration and unsteady discharge terms, the model can be applied for analyses of purely unsteady cases where the pump dynamic characteristics show considerable departure from their steady-state characteristics.

Delay Optimal Event Detection on Ad Hoc Wireless Sensor Networks

We consider a small extent sensor network for event detection, in which nodes periodically take samples and then contend over a random access network to transmit their measurement packets to the fusion center. We consider two procedures at the fusion center for processing the measurements. The Bayesian setting, is assumed, that is, the fusion center has a prior distribution on the change time. In the first procedure, the decision algorithm at the fusion center is network-oblivious and makes a decision only when a complete vector of measurements taken at a sampling instant is available. In the second procedure, the decision algorithm at the fusion center is network-aware and processes measurements as they arrive, but in a time-causal order. In this case, the decision statistic depends on the network delays, whereas in the network-oblivious case, the decision statistic does not. This yields a Bayesian change-detection problem with a trade-off between the random network delay and the decision delay that is, a higher sampling rate reduces the decision delay but increases the random access delay. Under periodic sampling, in the network-oblivious case, the structure of the optimal stopping rule is the same as that without the network, and the optimal change detection delay decouples into the network delay and the optimal decision delay without the network. In the network-aware case, the optimal stopping problem is analyzed as a partially observable Markov decision process, in which the states of the queues and delays in the network need to be maintained. A sufficient decision statistic is the network state and the posterior probability of change having occurred, given the measurements received and the state of the network. The optimal regimes are studied using simulation.

Existence of Value in Stochastic Differential Games of Mixed Type

In this article, we address stochastic differential games of mixed type with both control and stopping times. Under standard assumptions, we show that the value of the game can be characterized as the unique viscosity solution of corresponding Hamilton-Jacobi-Isaacs (HJI) variational inequalities.