The fact that treatment with cytotoxic agents can induce myelodysplasia (MDS) and acute myeloid leukaemia (AML) in patients suffering from cancer and from other diseases has now been thoroughly documented (Kyle et al., 1970;Reimer et al., 1977;Casciato & Scott, 1979;Berk et al., 1981;Coltman, 1982;Green et al., 1982;Pedersen-Bjergaard & Larsen, 1982;Boice et al., 1983;Boivin & Hutchinson, 1984;Lakhani, 1984). However, quantitative relationships with the dose and duration of treatment for different agents, and the relative oncogenic potential of different drugs have yet to be clearly elucidated. It is also not known whether susceptibility to this outcome is influenced by the condition being treated. As a class, the alkylating agents have been shown to be particularly capable of inducing MDS and AML, and there are many reports of both cyclophosphamide and melphalan inducing MDS and AML in patients with mylelomatosis and other neoplasms (Kyle et al., 1970;Gonzalez et al., 1977;Bergsagel et al., 1979;Casciato & Scott, 1979;Buckman et al., 1982;Coltman, 1982). A quantitative comparison of the relative potency in doing so of these two drugs is more difficult to obtain. In the first two trials in myelomatosis of the Medical Research Council's Working Party on Leukaemia in Adults patients were allocated at random to be treated by either cyclophosphamide or melphalan, and the long-term follow-up of these patients offers a rare opportunity to assess the relative oncogenic potential of these two drugs. Patients and methodsThe population at risk comprised all patients randomised in the first two Medical Research Council trials (MRC, 1973; MRC, 1980a Evidence of MDS or AML was sought from routine follow-up reports to the trial headquarters. In addition the clinical notes of all five-year survivors and all patients developing MDS or AML were used to abstract details of all chemotherapy, including specific drugs and combinations and the exact sequences of times on and off therapy. The diagnosis of MDS or AML was reviewed by one of us (D.A.G.G.) and confirmed where possible by review of blood films and bone marrow aspirates. Bone marrow investigations were performed only on patients who developed unexplained cytopaenia.In examining the blood and bone-marrow films, the diagnosis of MDS or AML was made according to the proposals of the French-American-British (FAB) cooperative group (Bennett et al., 1982). In brief, the diagnosis of leukaemia was made only when blast cells accounted for >30% of the nucleated cells in the bone marrow; when the blast-cell count was 5-20% the condition was diagnosed as refractory anaemia with excess of blasts (RAEB) and when it was .20% but <30% as RAEB in transformation (RAEBt). Myeloma cells were omitted in performing the differential counts. The slides from one patient from the first MRC trial had been examined for our first follow-up study (Buckman et al., 1982) and the diagnosis was of MDS and AML, but the slides were not available for review on this occasion. Slides were available for review...

The methodology of Support Vector Machine Methods is adapted in a straightforward manner to enable the analysis of stratified outcomes such as horseracing results. As the strength of the Support Vector Machine approach lies in its apparent ability to produce generalisable models when the dimensionality of the inputs is large relative to the the number of observations, such a methodology would appear to be particularly appropriate in the horseracing context, where often the number of input variables deemed as being potentially relevant can be difficult to reconcile with the scarcity of relevant race results. The methods are applied to a relatively small (200 races in-sample) sample of Australian racing data and tested on 100 races out-of-sample with promising results, especially considering the relatively large number (12) of input variables used.

Abstract. The Bacterial Foraging Optimization (BFO) algorithm is a biologically inspired computation technique which is based on mimicking the foraging behavior of E.coli bacteria. This paper illustrates how a BFO algorithm can be constructed and applied to solve parameter estimation of a EGARCH-M model which is then used for calibration of a volatility option pricing model. The results from the algorithm are shown to be robust and extendable, suggesting the potential of applying the BFO for financial modeling.

No abstract

No abstract